The AI IPO tsunami on the stock market has only recently gotten under way, with SpaceX’s more-than-$2 trillion IPO likely to be followed in several months by OpenAI’s and Anthropic’s IPOs — each of which is likely to hit $1 trillion.

That will mint three new trillion-dollar AI companies in a matter of months, all of which compete with Microsoft. 

Wall Street has never seen anything like it. Previously, the most money raised by all IPOs in a single year was $671 billion in 2021. It took 38,644 deals to get to that figure. Compare that to three deals this year that by themselves will likely total $4 trillion.

The numbers are eye-popping. 

For Microsoft though, it’s not the numbers themselves that are important. It’s what will happen to the company once it as three newly minted trillion-dollar AI competitors. Until recently, when it came to AI, Microsoft was king of the hill. But can it keep that place?

Microsoft’s weakened position

The IPOs come at a particularly fraught time for Microsoft. At one point, it was the most valuable AI company in the world, with a big head start on the tech industry.

No longer. Microsoft’s stock price has tanked in the past 12 months, even as the S&P index has soared. In the last year, Microsoft’s stock price has dropped 24%, while the S&P has jumped 24%. The two were pretty much in sync until last fall..

Microsoft has fared even worse against its most powerful AI competitor, Google. In June 2025, Google’s total value was $2.1 trillion, well behind Microsoft’s  $3.57 trillion. By  mid-June this year Google was worth $4.5 trillion, Microsoft, $2.8 trillion. That’s entirely due to Google’s AI push and Microsoft’s failure to improve Copilot significantly.

Matt Vellosso, who worked for Microsoft for 14 years — including four as technical advisor to CEO Satya Nadella and then as Partner Director for fostering AI innovation in Windows before leaving in 2023 — is scathing about what he views as Microsoft’s AI failures. 

He warned: “Microsoft missed the internet wave, the mobile wave and now it missed the AI wave.”

Show me the money

The most immediate likely impact on Microsoft after the three massive IPOs take place will be on the company’s stock price, which could well take another hit. Investors don’t want to miss out on the AI boom, and until now, there’s been a relatively small number of companies in which they could. When Anthropic and OpenAI join SpaceX as publicly traded companies, investors will have three more choices than they did only a few months ago. That could make it tougher for Microsoft to attract AI-focused investors. And that, in turn, could knock down its stock price. 

A lower stock price means the company will have a harder time raising money when it needs it. In addition, Microsoft won’t be able to as easily buy other companies in stock-only deals, because the value of its stock will be less.

Still, the IPOs are not all bad news for Microsoft’s stock position. It does, after all, own 27% of OpenAI. So if OpenAI is valued at $1 trillion after its IPO, Microsoft has a $270 billion stake in it.

Thumbs up for increased Azure revenue

One area where these mega IPOs should be unalloyed good news for Microsoft is in Azure revenue, which should significantly. Microsoft’s final divorce settlement with OpenAI requires the latter to buy $250 billion in Azure services through 2030. And its  estimated trillion-dollar valuation ensures the company will be around for the long term, meaning Microsoft can count on that revenue — and possibly more on an ongoing basis.

Microsoft will likely also get tens of billions of dollars from Anthropic for Azure cloud services. Anthropic might spend $43 billion annually on Azure cloud services by 2030, according to estimates by HSBC, one of the world’s largest banking and financial services companies.

Thumbs down for the effect on Copilot

Although Azure use will boom thanks to the IPOs, Microsoft’s own Copilot could face significant competition at a time when it’s having serious problems gaining traction. In its April 2026 earnings call, the company said Microsoft 365 Copilot had 20 million paid seats. That’s up from 15 million paid seats as of January, but still a minuscule number, considering Microsoft has 450 million Microsoft 365 commercial subscribers. That means only about 4% of the company’s business customers have been willing to pay for it since its launch in late 2023.

Beyond that, developers have been leaving GitHub Copilot and turning to Anthropic’s Claude Code and SpaceX’s recently purchased Cursor. A survey by the software development company JetBrains of more than 10,000 developers found that 29% used GitHub Copilot, 18% used Cursor and 18% used Claude Code. 

Early in 2025, GitHub Copilot had a commanding 67% market share.

More downside than up from the IPOs

Although the IPOs do have some upside for Microsoft — dramatically increased Azure revenue and several hundred billion dollars in OpenAI stock holdings — mostly, they’re bad news thanks to increased competition. 

Unless Microsoft significantly improves Copilot for businesses, developers and consumers, the company’s one-time AI dominance will erode even further.

AI has created a tough job environment for entry-level workers and things aren’t getting better anytime soon — even those with AI capabilities now need “senior-level” skills to land a job.

“AI-exposed entry-level roles are seven times more likely to require traditionally senior-level skills such as judgement and leadership,” consulting firm PwC said in a study released this month.

That’s because AI is changing the traditional career ladder. Companies are increasingly looking for candidates that use the cutting-edge tools and services to amplify their performance and grow faster. “Organizations must rethink how they mentor and train junior staff, helping them step up to complex decision-making much earlier in their careers,” PwC said.

Entry-level job seekers with or without AI skills are already dealing with stagnant wages,  layoffs, and stalled hiring. 

(The PwC findings echo similar concerns raised late last year in McKinsey’s State of AI report. Many companies are reducing headcount by deploying AI agents to take over entry-level jobs.)

Early-career AI job postings “have flatlined in highly AI exposed sectors,” and listings for junior roles with mid-career or senior-level skills have grown 35% since 2019, PwC said.  The consulting firm largely discounted the notion that AI is taking jobs away, though other studies point in the opposite direction. 

By the end of May, AI-driven job cuts had reached 87,174 for 2026, already outpacing the total of around 54,836 in 2025, according to figures released by Challenger, Gray and Christmas earlier this month.

The AI-driven layoffs haven’t reached the “jobpocalypse” stage yet, and workers are more productive with it, said Andy Challenger, chief revenue officer at Challenger, Gray and Christmas. But companies are rethinking hiring and long-term operational strategies as AI becomes a routine component in daily workflows and processes, he said.

Businesses are “restructuring aggressively as they reposition for an AI-driven economy,” he said.

That’s putting downward pressure on entry-level hiring as AI tools absorb more routine work, said Kye Mitchell, head of Experis US, a part of ManpowerGroup. “That doesn’t remove opportunity, but it changes the expectations. Employers now expect candidates to come in with hands-on experience, AI familiarity, and the ability to contribute faster,” Mitchell said.

Compensation remains strong for specialized, in-demand skills, while more commoditized roles such as customer service, helpdesk, and some entry-level positions  are flattening. “The shift overall is toward skills-based hiring, where demonstrable capability matters more than credentials alone,” Mitchell said.

Graduates who combine technical fundamentals with practical experience, AI fluency and strong communication skills stand out quickly. Job candidates can’t rely solely on academic credentials. 

“Employers are moving away from ‘train-from-scratch’ hiring and looking for talent that can contribute earlier and continue to adapt,” Mitchell said.

The PwC study also focused on the productivity gap between companies that have invested heavily in AI and companies lagging in adoption.

Since ChatGPT showed up in 2022, AI-exposed companies have seen productivity gains of 40% versus other companies. “The companies achieving the biggest productivity gains from AI are not using it only to cut costs,” PwC said.

AI-forward firms are also raising headcounts and wages. “Far from being a job killer, AI may actually be a job expander when used to unlock growth and enter new markets,” PwC said. 

Workers who use their domain expertise to supplement AI tools can advance, with AI-exposed roles “2.5 times more likely to rely on skills like empathy, judgement, and creativity that become even more valuable as AI absorbs some routine work,” PwC said.

Microsoft has been pushing hard to make Visual Studio Code a major way to consume its AI services, mostly in the form of GitHub Copilot. GitHub Copilot’s deep integration with VS Code brings many conveniences — inline autocomplete, for instance — but it’s frustrating for those, like me, who would rather use another model provider, or even a locally hosted LLM, for those functions.

Visual Studio Code 1.122 introduced a new feature, “Use BYOK [Bring Your Own Key] without a GitHub sign-in,” that allows you to “use chat, tools, and MCP servers in air-gapped or restricted environments where GitHub sign-in isn’t possible.” More importantly, it “enables fully offline workflows with local models like Ollama.”

In other words, you can now use locally hosted LLMs for chat, tools, and Model Context Protocol servers inside Visual Studio Code. The one thing you still can’t do is use a local LLM for inline and next-edit suggestions — at least, not without additional tooling.

Choosing a model for BYOK mode

If you want to use a local LLM with VS Code’s bring-your-own-model system, the first thing you need is a way to host the model. VS Code lacks a model-hosting mechanism of its own, although it’s conceivable that a VS Code extension may offer something like that in the future. That said, hosting models is complicated enough that a dedicated app is really needed for the job.

One easy way to host models is via a product like LM Studio, a convenient GUI for standing up, serving, and managing LLMs on one’s own hardware. The model host does not have to be the same system you run VS Code on, either. It can be on a server box you control, or on a cloud instance.

The choice of model is also important. Many models are powerful but won’t run well on commodity hardware because they’re simply too big. A good rule of thumb is to choose a model that fits into existing VRAM, along with the memory needed for a sizable token context (the more, the better). Also, the model should be suited to coding and development work. Some models in this vein that fit comfortably into 8GB VRAM include:

• Gemma4 (effective 2 billion parameters version)
• Qwen3.5 9B
• Codestral 22B v.0.1 (proprietary license)

Setting up BYOK mode in VS Code

Once you have a model up and running, you can integrate it with Visual Studio Code. If you’ve disabled VS Code’s AI features, you will need to turn them on. Make sure the setting chat.disableAIFeatures is turned off. You can find it in Settings | Chat | Miscellaneous.

Third-party language models are managed through Visual Studio Code’s language model list. Press Ctrl-Shift-P and type Manage Language Models to open the list of existing language models.

First you will see a list of the built-in models, which are all externally hosted. To add a new model, select Add Models at the top right and select Custom Endpoint.

You’ll then get a series of prompts:

• Group Name: This is “Custom Endpoint” by default, but you can choose any name you want. The name is strictly for organizing the model list and doesn’t affect things like model recognition or connectivity.
• API Key: If you’ve configured LM Studio to use an API key for serving models, provide it here. If you’re hosting the model locally and you haven’t explicitly set up API keys, you can leave this blank.
• API Type: The options here are Chat Completions, Responses, and Messages. Most of the time you’ll want to use Responses, as it’s the most general-purpose option of the three.

Once you finish providing those answers, you’ll be dropped into a modal editor for a JSON file that holds the details about the endpoint you’re configuring.

You’ll need to provide a few more details by typing them into the labeled fields:

• id: A text field that uniquely identifies this particular entry. The choice of ID is pretty much arbitrary; if you’re using only a single model, the ID could be the model name.
• name: The name of the model that is used to identify it on the model server. In LM Studio, you can get this name by clicking on My Models in the main interface, then selecting the three-dot icon for the model in question and clicking Copy Default Identifier. For Qwen 2.5, for instance, name might be something like qwen2.5-coder-7b-instruct.
• url: The URL to the server’s endpoint. On LM Studio, this defaults to something like http://127.0.0.1:1234/v1. The /v1 at the end is important because that endpoint is used for autodiscovery of models and their capabilities.

The other fields generally don’t need editing. Most models have tool calling functionality. If you know for a fact that the model you’re using doesn’t have vision support, then set vision to false.

Once you have these fields filled in, you can close the modal editor to save the changes. If you reload the Manage Language Models page, you’ll now see your new endpoint:

You should now be able to launch the chat window and use the defined model for conversation and utilities:

One current, and major, limitation of Visual Studio Code’s BYOK functionality is that it only works for chat and utility tasks. It doesn’t allow you to use a local model for inline suggestions or code completions. The only way to take advantage of local models for expanded functionality with VS Code is to use a third-party tool like Continue.

It isn’t clear if Microsoft will eventually lift this restriction. GitHub Copilot integration in VS Code is a large part of how Copilot as a service reaches its target audience. For the time being, you can certainly use third-party and local models for a significant part of your AI-assisted development work in VS Code, and you can close the functionality gap with additional tooling.

Unless you’ve been living under an old woodpile in your backyard, you have certainly seen how agentic coding is rocking the software development world. Things are happening fast and furious, and keeping up is practically a full-time job. 

The latest area that is catching the attention of developers is how agentic coding is affecting the open source community. The open source movement has been defending the rights of folks to use, change, and contribute to software for many years. And of course, agentic coding is starting to become part of that process.

On the one hand, maintainers of open source projects rightfully are frustrated as they become overwhelmed with pull requests of dubious quality and usefulness being submitted by coding agents. On the other hand, as David Heinemeier Hansson notes, maintainers are starting to get a little snooty about accepting AI-written code, viewing it as somehow not worthy of being included. Some organizations have explicitly banned AI-generated submissions.

I get that they don’t want AI slop overwhelming their input queues. But I think it is a huge mistake to ban AI-written code outright.

Whose code?

Before I dig deeper into that notion, it’s important to look at another issue that arises from all of this: Who actually owns the code that AI writes? 

Copyright requires that a human produce the thing being copyrighted. If you prompt Claude Code with “Write me a CMS system” and then Claude writes you a CMS system that you check into a public GitHub repository unchanged, it’s not quite clear if that code is protected by copyright. However, if you prompt Claude Code with a specification and guidelines and then you work with Claude to refine the initial result, reviewing the code and making changes as part of an iterative process, then it could be argued that a human did produce that code. But it is not at all clear-cut legally. (Please note that I am not a lawyer.)

The current thinking is that the result of accepting verbatim the output of a simple prompt is not copyrightable, and that no one actually owns the code — an interesting notion in and of itself. 

But then the ethical question comes into play. If I find a bug in an open source project, I ask GitHub Copilot to fix it, and Copilot writes a clever and effective fix, then who cares who owns the code? Should a maintainer of the project reject such a pull request just because it was AI-generated? That seems silly to me, yet it is happening today. 

Our code

There is, too, the issue of license compliance for AI-generated code. As a general rule, LLMs generate code rather than copying it. They don’t copy and paste code directly from repositories. However, there have been cases where AI-produced code has resembled open source code so closely that the claim could be made that it is a copy. If this happens with GPL code, it could be a violation of the license to use it without the receiving code base being “infected.” Open source maintainers naturally should be concerned about this happening.

In the end, an open source maintainer should care about the quality and license compliance of submissions, not how those submissions were derived. Gatekeeping based on the source of code doesn’t seem like a good path towards project success. Good code is good code, no matter where it comes from.

Agentic coding is here, and the open source community needs to realize — and embrace — that inevitability.

The US government decision to force Anthropic to close down its latest and greatest AI models, Fable 5 and Mythos 5, was only the next step in a burgeoning battle between AI providers and ignorant politicians. 

Anthropic was on top of the world. Its Mythos 5 LLM had everyone excited. (If you believe the hype, it was kind of scary, too.) Even Anthropic CEO Dario Amodei admitted — or boasted? — that Mythos would bring an “enormous increase in the amount of vulnerabilities, in the amount of breaches” to us all. But with that fear came the promise of more AI power than ever. 

Then, the roof caved in.

On June 12, the US Commerce Department used its export-control powers to demand that Anthropic cut off access to its Fable 5 and Mythos 5 models for all foreign nationals, citing national security concerns and fears of jailbreaks. After figuring out it had no way to do that, Anthropic pulled both of its newest frontier AI models offline worldwide.

Just what an AI company needs! All other Claude models, like the Opus and Sonnet series, remain online. But, come on, AI sales are all about the newest and most powerful models.  

Adding insult to injury — and this is true at many high-tech companies — Anthropic has many employees who aren’t US citizens. This means Anthropic’s own programmers can’t work on their latest models.

Of course, this isn’t the first time US President Donald J. Trump and company have tried to put a spoke in Anthropic’s wheels. Back in February, Anthropic refused to give Defense Secretary Pete Hegseth the power to use its models to spy on American citizens and to power autonomous weapons.

This go-around, it wasn’t because Anthropic refused to kowtow to Trump’s officials. It was, they say, out of fear that these new models could be used to attack American interests. 

Mind you, no one in Trump’s regime has the tech chops to know just how dangerous, or not, any AI model is. As I recently noted, Trump’s AI executive order has no teeth. Nor, more to the point, is there anyone in the administration with a clue about AI.

Specifically, at the Department of Commerce, neither Commerce Secretary Howard Lutnick nor William Kimmitt, undersecretary of commerce for international trade, knows a thing about AI. Just what we need, more political hacks deciding tech policy. 

So, how did they discover that Fable and Mythos were theoretically a danger to the US? Good question. According to The Wall Street Journal, it was Amazon CEO Andy Jassy who told the Trump administration that Fable was untrustworthy. Guess what? AWS offers its own full cloud AI stack, starting with Amazon Bedrock for foundation models, Amazon SageMaker for training and deployment, and a growing set of agentic AI tools and services. 

In other words, Amazon is not a neutral party; it’s a competitor. Funny, that.

Sure, Amazon also has partnerships with Anthropic. But, in case you haven’t noticed, all the big AI companies are in bed with each other. That doesn’t stop them from fighting. What’s heating this up is that the AI companies are no longer offering flat-rate subscriptions and are replacing them with far more expensive, token-based pricing schemes. 

Armed with this information, Commerce gave Anthropic 90 minutes to fix its “problem.” Right. AI development is fast, but it’s not that fast. In addition, according to Anthropic, officials haven’t spelled out exactly what’s wrong. They only know that Commerce claims there was a “narrow, non‑universal jailbreak” in Fable.   

That’s it. That’s all. 

Anthropic has also observed, with reason, that similar jailbreaks are possible on other leading models, like OpenAI’s GPT‑5.5. Those others, however, haven’t been hammered with comparable export controls.

The AI and security experts who do have an AI clue believe Commerce is behaving stupidly. (You won’t get any argument there from me.) For example, in an open letter, “On Transparent AI Cyber Protections,” they said Commerce’s directive “has taken the best models away from defenders, created market uncertainty, and risked America’s AI leadership without any real risk to justify it,” warning that pulling capabilities “away from defenders without a good reason when our adversaries are rapidly advancing is dangerous.”

Exactly so. 

Besides, as Alex Zenla, co-founder and CTO of security company Edera, observed, Fable’s capability to identify insecure code sections is the baseline for any model you’d trust to write secure code. The same capability exists in GPT-5.5, Opus, Sonnet, and Kimi 2.7 — it’s not unique to Fable. Pulling Fable from defenders doesn’t remove the capability from the threat landscape. It just removes it from the people trying to build safer systems.”

This is not about AI safety or security. 

What this is really about is opening a new front in Trump’s war against the liberal-leaning Anthropic. Mind you, Anthropic isn’t really liberal. This has more to do with Anthropic not following in other tech firms’ groveling to Trump. 

However, Trump doesn’t seem to realize that by essentially shutting down Anthropic’s biggest move to date, he’s also telling the world that they can’t rely on American AI companies down the road. Sure, in the short run, this hurts Anthropic. In the long run, it’s going to be another reason for Europe and other countries — taking digital sovereignty seriously — to avoid doing business with any American tech company. 

In the meantime, Anthropic and Trump administration officials are in tense talks over whether, and under what safeguards, the models could return to the marketplace. Commerce indicates it might allow a narrower relaunch if jailbreak issues are resolved and additional controls are in place — whatever those might be. 

Since it’s really all about massaging Trump’s ego, I’m not feeling terribly optimistic. Just ask Israeli Prime Minister Benjamin Netanyahu how well letting Trump set the terms of engagement goes. It’s not pretty. 

Nvidia and Microsoft this month touted the reinvention of computers with a new class of “agentic AI PCs” that will “reinvent the way PCs work.” 

That’s how Nvidia CEO Jensen Huang described the computers at the recent Computex trade show. At the event, Nvidia introduced its first AI-focused PC chip called N1X, which has an integrated CPU and GPU and will be used in agentic AI PCs.

Nvidia’s new RTX Spark PCs are the first in a major “PC reinvention for 40 years,” Huang said, likening them to AI phones. “You could talk to it, it could look at you. You could ask it to read files… [or] go help you do research.”

Not so fast, say analysts, who argue the computers are mostly repackaged AI PCs that shouldn’t necessarily drive enterprise upgrades.

“Agentic AI PCs is a strange term that should probably be deemphasized,” said Leonard Lee, principal analyst at neXt Curve. “Depending on use case, PCs of the last two generations are ‘agentic AI’-capable.”

Skeptical of the hype, analysts said many current PCs are already capable of running agents, such as those spun up on device by OpenClaw. Those agents don’t need to communicate with AI models in the cloud.

In fact, Lee said, Apple has already shown that Windows isn’t needed to run agentic workloads, with the Mac Mini among the most talked-about device for hosting personal AI agents and OpenClaw.

Analysts said the definition of agentic AI PCs really comes down to the hardware inside, which largely defines their capabilities. Most are really just an evolution of earlier AI PCs, which arrived a few years ago amid similar hype.

AI PCs, specifically CoPilot+ PCs, were originally designed with neural processing units (NPUs) to support the requirements of Windows Recall. “Some argue agentic AI PCs need more GPU compute, but…Qualcomm and Microsoft would counter that agentic AI PCs have been around for a while,” Lee said.

To highlight the capabilities of the RTX Spark PC, a demonstration at Computex showed how an architectural design workflow could be split between the PC and the cloud. (An MCP server managed data exchange to facilitate the cloud-PC workflow.)

Adobe has reengineered Photoshop and Premiere for RTX Spark PCs, allowing the software to communicate with AI agents on PCs and tools to run twice as fast, Huang said.

And at the company’s recent Build conference, Microsoft CEO Satya Nadella demonstrated the new Surface Laptop Ultra AI PC based on the RTX Spark design. He positioned agentic AI PCs as running a new execution layer that can act across files and devices, including generating and executing code. 

While AI PCs with agents will be capable, they still lack applications, said Jim McGregor, principal analyst at Tirias Research. “However, PCs — especially the small form factor PCs and workstations — are likely to be the home AI appliance,” he said.

For enterprises, upgrading to RTX Spark PCs at this point could pose problems. That’s because Nvidia’s N1X CPU is based on the Arm processor architecture and could run into compatibility issues with x86 applications designed for Intel and AMD chips.

“Despite Arm processors being available for a while, enterprise penetration stays small, since compatibility with all apps, drivers, and corporate systems is critical and requires extensive testing…, even though Microsoft has done a reasonably good job with Windows 11 on Arm,” said Jack Gold, principal analyst at J. Gold Associates.

Because Microsoft is moving to natively incorporate AI into Windows, those PCs will land in enterprises ultimately, Gold said.

Regardless of hype, it might be a good time for enterprises chasing an agentic AI agenda to upgrade laggard Windows 10 devices, Lee said. “Considering the considerable hype about agentic AI, early adopters will need to upgrade their fleets if they pursue a broad agentic AI agenda for their organizations,” he said.

In the past year, the enterprise AI ecosystem has gained enormous capability and zero consensus.

Developers now have a remarkable set of tools for building AI agents: OpenAI’s frameworks, Anthropic’s Claude tooling, LangChain, LangGraph, CrewAI, Microsoft AutoGen, and a growing list of alternatives. Each promises to coordinate reasoning loops, manage multi-step task execution, and connect agents to tools and APIs. For experimentation, the progress has been substantial. Teams can now assemble sophisticated agent workflows in days that would have taken months two years ago.

But I’ve watched this pattern before. In over two decades of building and selling distributed systems platforms, I’ve seen the same dynamic play out across nearly every major infrastructure shift: the tools for consuming a new capability arrive before the infrastructure for governing it does. The gap that emerges isn’t immediately obvious in development environments. It becomes obvious in production.

That’s exactly where enterprise AI stands today.

What agent frameworks don’t handle

Modern agent frameworks are fundamentally coordination systems. They determine what a system should do: which tools to call, how to sequence tasks, how to delegate work across agents. That’s hard work, and they’ve gotten quite good at it.

What they rarely address is where those tasks are allowed to run, and under what conditions.

Take a seemingly simple workflow: summarize customer support transcripts using an LLM. In a development environment, the implementation is clean. The agent calls a model API, passes the transcript, and returns a summary. In production at an enterprise, the same request may involve a dataset that can’t cross a specific geographic boundary, a model that isn’t approved for regulated data, and an audit requirement that demands a traceable record of what happened.

Those aren’t planning problems the agent framework was designed to solve. They’re execution governance problems. Most frameworks quietly assume they’re handled somewhere else in the stack. In many enterprise environments, they’re not handled at all. Gartner predicts more than 40% of agentic AI projects will be canceled by the end of 2027, citing inadequate risk controls as a primary driver of failure—a number that reflects exactly this gap.

What the missing layer actually does

Addressing these governance problems requires an additional layer between agent logic and execution: one that evaluates every agent action against policies governing where data can reside, which models may process it, who authorized the request, and how the action fits within the organizational context. The agent framework determines what the system should do. The orchestration layer determines whether and where it’s allowed to happen. Keeping those responsibilities separate allows both layers to evolve independently. It also means you can adopt new agent frameworks without rebuilding your governance model from scratch.

This separation will feel familiar to anyone who has worked through the Kubernetes era. Kubernetes doesn’t care what’s inside your container. It finds capacity, allocates resources, and ensures things run. The orchestration layer for agentic AI plays an analogous role: it doesn’t care which agent framework generated the request. It enforces the conditions under which that request can execute.

Richer authorization models

Traditional enterprise access control is built around a simple question: can user X access resource Y? That’s insufficient for autonomous agents.

A realistic authorization decision for an agent request might look more like this:

request = {
    "agent": "support-summary-agent",
    "task": "summarize",
    "dataset": "customer_support_logs",
    "model": "external_llm_api",
    "delegated_by": "user_4821"
}

policy = evaluate_policy(request)

if policy.allowed:
    route_to_execution(policy.execution_environment)
else:
    raise AuthorizationError(policy.reason)

The policy engine here evaluates dataset classification, model approval status, geographic processing rules, and the delegation chain that initiated the request. That might mean redirecting the task to an internal inference cluster instead of a public API endpoint, or blocking the request if no compliant execution environment exists. From the agent’s perspective, the task still executes. The orchestration layer ensures it runs in an environment that satisfies enterprise policy.

Why ontologies are load-bearing infrastructure

For the orchestration layer to make good decisions, it needs to do more than label data. It needs to understand how the entities involved in a request relate to each other, and reason over those relationships to determine what’s allowed.

Consider the customer support transcript example again. Metadata tells you the dataset contains PII (personally identifiable information). An ontology lets the system reason across a connected chain: the task operates on a dataset containing personal data; that data is governed by GDPR; the organization’s policy requires processing within an approved EU environment; the selected model runs outside that boundary. From those four connected facts, the orchestration layer can infer the request must be rerouted or blocked. The system reasoned over the relationships rather than matching against a hardcoded rule tied to a specific dataset.

This is what makes policy enforcement, execution routing, data locality, and audit decisions computable at runtime. An ontology can be built around virtually any entity-relationship set the enterprise needs to govern: datasets, models, agents, users, regulations, tasks, environments. The relationships that matter are the ones that drive the decisions the governance layer needs to make. Access control lists can restrict who touches a resource, but they can’t reason across a connected set of entities. That reasoning is what the orchestration layer depends on.

Decision provenance as a first-class requirement

Enterprise systems also require auditability. When automated agents trigger actions across multiple systems, organizations must be able to reconstruct the decision path that produced the outcome. Compliance depends on it. So does incident response and basic operational trust.

An orchestration layer generates records describing the initiating identity, the agent, the model, the data sources, the policies evaluated during authorization, and virtually anything else the organization chooses to capture in its ontology. That chain of custody allows teams to investigate incidents and validate compliance without treating production AI systems as operational black boxes.

Regulators and auditors are no longer satisfied with knowing what an AI system was designed to do. They want a factual record of what it did in a specific instance, under what authorization, and with what effect—something dashboards can’t provide, but a well-designed orchestration layer can. The EU AI Act makes this explicit: under Article 12 and Article 17, high-risk AI systems must maintain documentation that makes decisions traceable and auditable, with records sufficient to support investigation after the fact.

Where this leaves enterprise teams

Agent frameworks will keep improving. The coordination problems they solve are real, and the ecosystem will continue to mature. But the architectural challenge for enterprises has shifted. It’s no longer primarily about coordinating agents. It’s about governing how those agents interact with real infrastructure, real data, and real compliance obligations.

The patterns for doing that exist today: contextual authorization, data locality enforcement, ontology-aware policy evaluation, decision provenance. What most organizations are missing is the recognition that these capabilities belong in a distinct layer that operates independently of whichever agent framework sits above it. Build that layer, and the rest becomes manageable.

—

New Tech Forum provides a venue for technology leaders—including vendors and other outside contributors—to explore and discuss emerging enterprise technology in unprecedented depth and breadth. The selection is subjective, based on our pick of the technologies we believe to be important and of greatest interest to InfoWorld readers. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Send all inquiries to doug_dineley@foundryco.com.

Tech industry experts are urging IT decision-makers to be wary of AI vendor gimmicks such as free tokens, and to adopt a multi-vendor and multi-model strategy to avoid vendor lock-in.

“Don’t be afraid to adopt a multi-vendor approach to get value from different AI tools rather than risk lock-in with a single one,” said Max Goss, senior director analyst at Gartner.

It is unlikely one AI vendor or model will meet an organization’s requirements, Goss said.

The advice comes as more AI vendors are offering cheap tokens subsidized by venture capital in a land grab for customers. The companies are also hiring forward-deployed engineers (FDEs) to push their models to enterprises.

Once companies start developing business processes around specific AI models, they get locked into their ecosystem. “People are adopting hybrid strategies…to cut token costs, and adopting more token-efficient models,” said Jack Gold, principal analyst at J. Gold Associates.

Free and low-cost tokens from AI vendors could incentivize companies to build processes and workflows around proprietary LLMs and agents, said Max Leaming, head of data science and AI solutions at ManpowerGroup.

But as the AI landscape evolves, it’s difficult to predict whether a multi-model or multi-vendor landscape will emerge, said Logan Wolfe, partner of global AI strategy and sovereign transformation at IT consulting firm Kyndryl. “I think it could be multi-model, yes. It really comes down to the use case and the type of implementation that you’re having,” Wolfe said.

Enterprises are still in the midst of moving blue-sky experimentation to a mindset where they see AI as a powerful tool that needs to make sense from a business perspective. With that in mind, IT leaders should ground their AI strategies on use cases as opposed to vendors, Wolfe said.

“If it’s a highly regulated space, if it’s a financial sector, a healthcare sector, then you will be placing a lot more emphasis on safety, privacy, maintaining certain regulations, and so that could prevent you from rapid model switching based on cost,” Wolfe said.

For low-stakes use cases, it would be prudent to have a model-switching approach that doesn’t break the bank. “For a low-hanging fruit use case with varying volume, like a customer support data center, during heavy load times you could switch to the more capable model, then optimize that on evenings and weekends,” Wolfe said.

ServiceNow Chief Digital Information Officer Kellie Romack, who’s worked in IT for 25 years, said companies need to understand how their AI is built. “You can’t have AI built in such a way that you don’t have human beings understanding how it was built…, how to debug, back up, and retrace,” she said.

Romack has also long resisted ripping out one vendor’s platform to replace it with another. “I say, ‘Let’s talk about the technology you already have…, now let’s see the best of breed,’” she said.

After studying what customers already own and where their contracts and plans are headed, Romack looks at options based on architectural principles and the problem being solved, then runs multiple models in-house, such as Anthropic’s Claude and Microsoft’s Copilot, through one LLM gateway.

“We have a lot of different things in-house that people can put their fingers on,” Romack said.

For example, Claude might be better for reading a long Word document, while Copilot might be better for a quick summary.

She is sensitive about internal AI spending. “Every day we look at token spend. I’ll look at an engineer that’s got the same job as another engineer, and I’m like, ‘OK, you spent $10, you spent $10,000. Why?’”

Avoiding vendor lock-in is important for continuity of service. Outages hit AI services from OpenAI and Claude in recent months, and a multi-model approach provides fallback options, Gartner’s Goss said.

“If you are relying on a single provider with a single model, there’s risk there. You can mitigate that risk with a multi-model approach,” he said.

It was the best of trends, it was the worst of trends. It was the epoch of artificial intelligence,  it was the epoch of artificial inflation. 

I’m truly excited to be alive at a time when AI, formerly the stuff of science fiction, is now an everyday reality and promising so many benefits to humankind. 

AI is accelerating drug discovery, slashing the cost of protein folding research, diagnosing cancers earlier than human radiologists can, automating the drudgery out of nearly every white-collar job, translating speech across hundreds of languages in real time, and giving the blind a way to see the world through a camera. 

AI is giving us all this and so much more. But no amount of techno-optimism can hide the fact that AI is making just about everything more expensive. 

While the AI trend is making a tiny number of rich people even richer, the public at large is paying the price through rapidly rising prices; it represents a transfer of wealth from the have-nots to the haves. 

AI is a machine that eats resources. It eats chips. It eats electricity. It eats water, land, labor, and building materials. AI’s gluttony creates scarcity, and scarcity creates inflation. 

Here are all the ways AI is driving up costs for you and me. 

Your gadgets cost more

AI runs on memory chips and other computing hardware. AI companies buy so many of them that they’ve created a shortage. NAND prices shot up around 246% from the start of 2025 through last December, according to Kingston. Hard disk drive prices in Europe rose 46% in just four months. 

The chip shortage pushed smartphone prices to all-time highs — one analyst called it a “tsunami-like shock” to the industry. Beyond that, computer and device prices will climb by 20% by the end of 2026, according to one estimate.

Apple CEO Tim Cook told the Wall Street Journal this week that price increases on Apple products are “unavoidable,” citing the surging costs of memory and storage chips driven by AI data-center demand. He described the supply shock as a “hundred-year flood.”

Software costs more

Greedflation has hit the software industry. Salesforce, ServiceNow, and others have increased subscription prices, blaming AI for the hikes. IT spending globally is forecast to grow 13.5% in 2026 over the previous year, reaching $6.31 trillion, according to Gartner, with the bulk of the increase driven by AI infrastructure investments.

The pricing models themselves are getting more complex and more expensive: seat licenses plus API usage plus GPU compute plus data storage plus compliance layers. What used to be one subscription is now five line items.

Enterprise software spending is growing 13.3%, with much of it being price increases on existing contracts rather than new purchases.

Services cost more

The services you buy cost more because companies are burning money on AI and passing on the costs to you. While tokens are getting cheaper, the new reasoning models can use anywhere from several times to tens of times more tokens than traditional models for comparable tasks.

SaaS inflation now runs at 13.2%, which is nearly five times the consumer inflation rate, and a majority of that increase is due to AI costs. 

Goldman Sachs forecasts that agentic AI could drive a 24-fold increase in token consumption by 2030, and that applies to the companies that provide enterprise services. 

Electricity costs more

AI data centers draw power the way a city does, pushing US residential electricity prices up roughly 5% on an annual average basis in 2025. That’s nearly double the general inflation rate of 2.7%.

Wholesale electricity prices near data center clusters have more than doubled since 2020. Goldman Sachs says electricity inflation will hover around 6% through 2027. 

The utilities build new power plants and transmission lines to serve these data centers, then hand the bill to everyone on the grid. You pay for AI’s appetite, whether you use AI or not. 

Your heating bill goes up, too, because the same natural gas that warms your home is being burned to generate power for data centers.

Your car costs more

Modern cars are computers on wheels. And a new automotive chip shortage is now underway. Prices for the memory chips that go into cars are expected to rise 70% to 100% in 2026 adding as much as $400 to the price of a car. 

Your house costs more

Data centers need land near power lines and water. They buy it at record prices, and they outbid the people who would have built homes on it. In Texas, data centers compete directly with homebuilders for utility-ready lots. In Northern Virginia, the data center buildout is squeezing the housing supply in a market that is already short. In Columbus, Reno, and Salt Lake City, data center land deals are pushing up land values beyond anything those markets have ever seen for industrial property. And the houses that do get built cost more, because data center construction has driven up wages for workers by 25% to 30%. 

Data centers even compete with public infrastructure projects for the same crews and the same concrete, copper, and steel — driving up the cost of roads and public works.

Everything costs more

AI is helping all kinds of companies fleece customers. Research from Carnegie Mellon in 2025 found that AI-driven ranking and pricing systems raise prices. 

AI-powered pricing algorithms now set the price of your rideshare, your flight, your hotel room, even your concert ticket. They use AI to estimate the maximum amount you’re willing to pay, then charge you that amount. Called surge pricing or dynamic pricing, the bottom line is that it affects your bottom line. 

A 2020 study in the American Economic Review showed that AI algorithms create a poverty premium: They learn that people with fewer alternatives are less sensitive to price, so they charge poor people more. 

Here’s another weird phenomenon hardly anyone talks about: When competing companies all use similar AI pricing systems, they can arrive at higher prices together without ever talking to each other. It adds up to a kind of accidental price-fixing. 

Food costs more

Higher electricity prices flow through the entire economy. Farms, food processors, trucking companies, and stores all pay more for power because of AI consumption, and they pass those costs down the chain until they’re ultimately paid by food consumers. 

Data centers are also consuming land that was once used for farming, forcing some farms to locate further away from population centers. 

Taxes cost more

Data centers receive enormous tax incentives and subsidies from state and local governments. At least 38 states now offer such incentives to data centers, which means funding shortfalls have to be made up by families paying their taxes. 

Texas is projected to lose $3.3 billion by 2029. Meta’s got a 20-year sales tax exemption from the state of Louisiana on data center equipment worth an estimated $3.3 billion. Pennsylvania will probably give around $2 billion in data center tax breaks. 

AI may one day change the world. But for now, it’s mainly just changing the cost of living. 

AI disclosures: I don’t use AI for writing. The words you see here are mine. I used a few AI tools via Kagi Assistant (disclosure: my son works at Kagi) as well as both Kagi Search and Google Search as one part of my fact-checking for this column. I used a word processing product called Lex, which has AI tools, and after writing the column, I used Lex’s grammar checking tools to hunt for typos and errors and suggest word changes. Why I disclose my AI use and encourage you to do the same. 

A new form of vendor lock-in is here. And it’s not proprietary languages or rigid enterprise software suites — it’s something more fundamental. It’s the very thing that writes the code.

JetBrains Research found that 74% of developers worldwide use AI tools. Claude Code, available only since May 2025, is now the most popular AI coding tool, followed by Gemini Code Assist and GitHub Copilot, according to Jellyfish’s 2026 State of Engineering Management Report.

The latter study also found that 91% of developers say their productivity has increased in the past 12 months. As coding output expectations are rewritten daily, the engineering world is becoming heavily reliant on paid external AI services.

Gartner predicts that by 2028 spending on AI coding tokens could exceed developer salaries. Yet, tokenmaxxing while vibe coding through a vendor’s cloud-based API feels like a far cry from the open foundations of free programming languages and open models, which many of today’s AI platforms now abstract.

“Open infrastructure will be the backbone of the AI era,” says Peter Farkas, CEO of Percona, a provider of open-source database solutions. “Right now, too many companies are building their entire AI strategy on top of proprietary platforms because the convenience is seductive.”

“It’s ‘three clicks’ to stand up a database or an AI service in a hyperscaler, and that convenience blinds people to the lock-in they’re signing up for,” he adds. “As AI workloads mature, organizations will realize that depending on one vendor for their data, models, runtime, and pricing is not a strategy.”

AI-assisted coding is democratizing software engineering for non-engineers and accelerating top performers. But if teams are always working within the confines of how one platform thinks the world should work, it could create locked-in toolsets at scale. And as AI platform costs rise, a fundamental question arises: will software developers consume AI on their own terms, or on someone else’s?

There’s a strong case that the long-term winners in tech will be built on open-source standards and foundations, similar to the history of cloud-native computing and the internet itself.

“Open always wins,” says Brian Alvey, CTO at WordPress VIP, a managed WordPress hosting platform. “Not because it’s a fancy ideology, but because it gives you total freedom to adapt, evolve, and stay in control.”

Open infrastructure avoids a future where developers perpetually rent. “For AI to be useful to people at large, it can’t be something you’re paying rent for the rest of your life,” says Manik Surtani, CTO and co-founder of the Agentic AI Foundation (AAIF), a vendor-neutral home for open-source agentic AI technologies. “And it can’t be concentrated in one particular corporation or a small handful of corporations, because we know how that goes.”

Pricey, closed, proprietary AI

AI development today is traveling two parallel paths. On one path, open-source AI is thriving and fueling tremendous growth in the number and variety of AI models and tools. Just take the thousands of open-weight models on HuggingFace, the community around the OpenClaw AI agent, or the many academic institutions publishing new breakthroughs.

“Open-source models and tooling are hot on the heels of state-of-the-art, with interesting and boundary-pushing work being shared by labs and researchers across the world,” says Austin Parker, director of AI strategy at Honeycomb, an observability platform provider, citing frontier open-source models like Mistral, DeepSeek, and Ai2’s OLMo as examples.

Others agree. “There’s unprecedented openness at the model and tooling layer, with open-source models, frameworks, and orchestration advancing at remarkable speed,” says Mark Collier, general manager of AI and infrastructure at the Linux Foundation.

On the other path, we’re seeing heavy reliance on proprietary AI systems controlled by Anthropic, Cursor, Google, Microsoft, OpenAI, and others. As Collier says, “Many platforms are wrapping those open components in closed, opinionated interfaces that trade short-term speed for long-term constraints.”

Open source and the AI tooling market don’t always mix well. LangChain’s Open Agent Platform, for instance, was open-sourced to much fanfare in 2025, but by 2026 had been deprecated, with the repository now recommending fully managed alternatives.

For Roman Shaposhnik, co-founder and CTO of Ainekko, provider of an open-source, composable AI stack, the current AI platform landscape is reminiscent of low-code and no-code platforms, which promised democratization of software development but often failed to deliver, becoming synonymous with platform lock-in and inflexibility.

“Honestly, it feels familiar,” Shaposhnik says. “We have incredibly powerful AI tools right now, but most of them come bundled as tightly controlled platforms.” This is a risk for AI, he says, because the infrastructure, models, and hardware are tightly coupled. “If those layers are closed, you lose flexibility fast.”

Some abstractions that sit on top of models, like routing and agent frameworks, tend to be tightly coupled and optimized for certain models. Other platforms take the walled garden concept quite literally. Anthropic, for instance, has repeatedly made headlines for blocking access to its Claude models over vague policy violations. The company recently shut off competitor xAI’s use and stonewalled OpenCode, drawing community backlash.

Moves toward increasingly closed systems don’t bode well for an AI economy already built on shaky economics. As Vikram Srivats, head of product experience at WaveMaker, provider of an agentic application development platform, adds, “Given the unit economics of AI tooling and pace of accelerated change to keep up, it seems obvious that some will evolve to more of a closed system to be able to monetize and gain ROI.”

Why openness matters in the AI era

Reliance on proprietary AI platforms can create long-term operational dependencies. As systems become less interoperable, organizations may be forced to standardize on a single stack across data pipelines, models, and decision logic, says the Linux Foundation’s Collier.

“As infrastructure consolidates, enterprises become more exposed when platforms change direction, raise prices, or fall behind technically,” he says. “If you can’t change platforms without re-architecting your AI systems, you’ve already given up too much control.”

“When you build on someone else’s platform, you have to live by their rules and those rules always change,” adds WordPress VIP’s Alvey. “We’ve all seen this before, businesses wasting time and money building to serve Google, Facebook, YouTube, and the App Store, instead of building to serve their customers.”

Platform lock-in can also create direct business risk. As Ainekko’s Shaposhnik says, “It usually shows up as higher costs, fragile systems, and growing risk when it’s time to change direction.”

At Ainekko, an internal group called the AI Plumbers focuses on back-end AI infrastructure like inference, scheduling, memory, and hardware integration. “Their view is simple,” says Shaposhnik. “If those layers are closed, everything above them becomes fragile.”

Open standards, interfaces, and infrastructure provide a necessary hedge against closed systems to prevent this sort of fragility. “In the AI era, open infrastructure gives enterprises control, portability, and choice at exactly the time they need it most,” says Percona’s Farkas.

It can cost upwards of $100,000 to migrate enterprise software, according to Cloudaware, making portability a major enterprise concern. From this perspective, procuring closed systems can become a costly architectural dependency.

Others argue that openness is a critical hedge against vendor concentration risks at large, especially if AI replaces human labor en masse. “If all of that economic value is now being concentrated in the hands of one or two companies,” says the AAIF’s Surtani, “that’s an order of magnitude bigger problem than we’ve seen in any other wave of computing.”

Instead, open foundations allow adaptability to evolving conditions so enterprises can swap out models, agents, data, hardware, and orchestration, as needed. “Open standards let those components change independently without breaking the system,” says Collier. 

Openness can also help future-proof businesses against economic upheaval. “Open everything will help build a cushion for businesses and users to survive and thrive after the almost-certain correction in the current hype cycle,” says WaveMaker’s Srivats.

Momentum toward open AI infrastructure

At the industry level, momentum toward open AI infrastructure is growing. The establishment of the Agentic AI Foundation, Anthropic’s donation of Model Context Protocol (MCP), and Block’s donation of its Goose agent are significant ecosystem-wide moves toward openness. Other advances include the donation of llm-d, a Kubernetes framework for LLM inference, to the Cloud Native Computing Foundation (CNCF).

For Parker, donations like this help ensure long-term support and care. “Open standards aren’t just the foundation of the internet, they’re the foundation of the AI space,” he says. “I predict that we’ll see these practices continue, especially as enterprise adoption increases in earnest,” he adds.

Still, some question whether this level of stewardship is enough for a rapidly evolving ecosystem. “The internet benefited early on from groups that helped keep vendors aligned,” says Shaposhnik. “In AI infrastructure, we don’t really have that yet.”

“All of us open source veterans are hopeful,” he says, “but we also need to adapt to this new reality in what we do regarding AI infrastructure.”

Beyond industry governing bodies, companies themselves are also spearheading open AI initiatives. Warp, an agentic development environment, recently went open source amid closed-source rivals. Arcade.dev, meanwhile, is pushing an open-source Agent Library for agentic memory.

Where openness matters most in the AI stack

While AI infrastructure can be open in many ways, a few layers stand out as especially important. First is the openness of the model itself. “Open-source models must be the foundation of future trust and value,” says WaveMaker’s Srivats.

“The forms of open infrastructure that reduce integration friction and accelerate adoption stand out,” adds Neeraj Abhyankar, VP of data and AI at R Systems, a global digital solutions provider. For him, open model representation formats, open orchestration and execution layers, open agentic protocols, and open governance and metadata standards are all essential for enterprise flexibility.

Others place more value on the connective tissue between AI components. “The most important forms of open infrastructure are the ones that connect systems together,” says Collier. “That includes open APIs, metadata standards, identity and policy frameworks, and protocols for how models and agents communicate.” 

Arguably, MCP has become the connective tissue between AI agents and the broader API ecosystem. “If we get MCP right we unlock the same level of interoperability between entities on the web and models driving them as we came to enjoy during the Web 2.0 era and the API-first boom,” says Shaposhnik. “If we don’t we risk massive proprietary lock-ins.”

Parker agrees that open protocols will underlie future AI progress. “We’ll see continued development and progress on AI agents which will rely on protocols like MCP and ACP [Agent Client Protocol] to interoperate with various clients and each other,” he says. Yet a gap remains around API conventions for models. “It would be nice if we could get a commitment from model providers to use a standard here.”

For the AAIF’s Surtani, opening up the protocol layer is the most important aspect. “I think it’s really important for interoperability, for choice,” he says. “It means you can bring your own agent, you can bring your own framework, you can bring your own harness, and pick what model you want.”

Open standards may also play a significant role within inference architecture. “As AI expands to the edge, developers need visibility into how models run, how memory is used, and how performance scales,” says Shaposhnik. Open systems could make it easier to optimize, debug, and adapt while helping enterprises avoid observability fragmentation.

Lastly, cloud-native architectural standards are a key ingredient for open AI infrastructure. “We’re seeing Kubernetes become the missing link for people who want the hyperscaler-style convenience without hyperscaler lock-in,” says Percona’s Farkas. For him, Kubernetes has become the de facto hybrid enterprise deployment option for data, workloads, and AI components.

History repeats itself

The 2026 State of Open Source Report found avoiding vendor lock-in to be the primary driver of open source adoption. But beyond being a strategic decision for a single company, open infrastructure provides a layer for entire industries to be built upon.

Arguably, the internet itself is evidence of this, where groups like the IETF and the IEEE were instrumental in defining the fundamental protocols. “Without open protocols we would’ve been in telco hell and without phenomenons like Google or Facebook,” says Shaposhnik.

Or, take the history of Linux as a parallel. “Linux became the default operating system because it offered a common, vendor-neutral foundation that everyone could build on,” says Collier. “In the AI era, open infrastructure will define the layers that organizations rely on for long-term continuity.”

At the infrastructure level, open standards have repeatedly underpinned major platform shifts, from Docker to Kubernetes. The question now is whether AI will develop a similarly durable standards layer.

For Parker, it’s too early to say, but the current growth of AI mirrors the early cloud. “Remember that it took many years before we saw the development and popularization of the open source cloud-native ecosystem,” he says. “I think it would be a mistake to extrapolate from the current trajectory towards a closed, proprietary future.”

Others agree the future must be rooted in openness. “I see open infrastructure becoming the foundation of enterprise AI,” says R Systems’s Abhyankar. “As systems become more distributed and agent‑driven, closed ecosystems simply won’t scale.”

The groundwork is being laid through open agentic protocols, open frameworks, and industry support intended to reduce fragmentation around proprietary standards.

“Ironically, the AI movement has mostly seemed to learn from the mistakes of the past and is starting off on a more open foot,” says Parker. “Over time, I believe we’ll see innovation and openness thrive.”

As AI shifts from output-generating large language models (LLMs) to armies of agents taking actions on their own, there is a growing threat that failures could affect system reliability.

Temporal, a Bellevue, WA firm founded in 2019, hopes to solve that problem by stabilizing AI and long-running computing processes through “durable execution,” a technology that reliably resurrects failed computing processes on other hosts. If a machine crashes in the middle of agentic AI transactions, the company resurrects the function on a different host so it can continue exactly where it left off. Temporal says its durable execution process comes with a 100% durability guarantee.

The company was solving these kinds of distributed-systems problems long before the generative AI (genAI) rush, and today it powers infrastructure from Coinbase to Airbnb to OpenAI.

For IT decision-makers, Temporal’s offerings promise to bring reliability to AI operations — especially in regulated industries. Co-founder Samar Abbas spoke recently with Computerworld about his company’s technology, AI reliability, execution and what IT leaders need to keep in mind when evaluating and deploying agentic AI.

How did Temporal come to be a backbone powering AI system? “OpenAI recently announced we power many of their products underneath the cover — that’s where Temporal comes in.

“A couple of years ago, as LLMs got smarter, use cases were about generating outputs. Now, the industry is moving into agentic solutions, where AI isn’t just generating outputs, it’s taking actions. The longer these systems run, the more tools they invoke and the more value they create. 

“They’re evolving into a whole army of agents coordinating to get complex tasks done. This starts to look like distributed systems. We become that execution authority, the reliability backbone for those agentic systems to get their tasks done from A to Z.”

What is Temporal doing in the backend to keep them stable? “It’s a new paradigm for building applications, we call that durable execution. If you write a function and a failure happens, you lose all of your state. When a machine crashes, we maintain the state. We seamlessly resurrect that function on a different host and continue executing exactly where it left off, without you as a developer writing a single line of code. 

“It’s not just a cache, it’s an operating system guaranteeing execution in the presence of failure. We provide 100% durability guarantees behind it. People can use it for mission-critical use cases like transcribing doctor conversations, where even losing one [word] has big implications.”

Did this challenge exist before AI? “My co-founder and I have been at it pretty much our entire careers, spanning almost three decades. We’ve been solving this problem of distributed systems, at Amazon, Microsoft and then Uber, before starting Temporal. We are definitely pre-AI. 

“As more workloads move to cloud and become distributed, it introduces this class of failures. This is the fifth time I’m building this system. I built simple workflow, then [a] durable task framework at Azure, then Cadence at Uber, now Temporal. The company is roughly seven years old, but the code base is over 10 years old, because we started at Uber and it’s been running in production ever since.”

And now you tackle that same problem at AI scale? “The funny thing is, especially with AI, that same problem is now at such a massive scale, it’s gone on steroids, with agentic systems coming online and doing real work. We’ve been powering every Coinbase transaction, every Snap story, every Airbnb booking, every Yum Brands order —  KFC, Pizza Hut, Taco Bell.

“A couple of years ago, it looked like we’d over-invested in the scale and reliability of the platform. Then AI showed up in such a big way that now I’m asking the team: how do we scale another 10x, 100x, even 1000x? These agents are hitting the market in a way we couldn’t have imagined.”

Can you give a concrete example of that 100% guarantee? “When you’re moving money — debit or credit — imagine a failure happens after you’ve debited an account. Temporal guarantees execution of the entire function from start to finish, in the presence of all sorts of failures. At the end of the day, we’re in the business of selling reliability. 

“Developers today spend 80% of their time building resiliency into their systems by working with low-level primitives like queues, databases, retry mechanisms and durable timers, and stitching them together. What we provide eliminates that. And that reliability is exactly what’s becoming critical as enterprises move agents into production.”

Why should CIOs and other IT leaders care about AI reliability? “These agents are now real, no longer prototypes or [proofs of concept]. They’re doing meaningful work in those organizations. Reliability and durability weren’t on the radar for CIOs 12 months ago. Now, it’s the blocker. Especially in regulated industries, if a transaction gets left in the middle, it has real regulatory concerns. 

“That’s what’s holding them back from transitioning into the AI stage. It comes up in pretty much every conversation, in the first 10 minutes I’m talking to a CIO.”

CW: Where do the hardest problems sit today? “Security, governance, identity, authorization, authentication. This is where majority of the problems exist, and the ecosystem is so immature. Even things like MCP, I see organizations struggling to implement MCPs to build these agents. This is an area ripe for innovation and a lot of work, including here at Temporal.”

What skills should developers and CIOs build for the AI age? “The biggest skill now is that developers need to be more product focused. Product-minded engineers are the ones who will thrive, because the whole software development flow, writing code by hand, debugging by hand, testing it, is going to look completely different. The job of an engineer is now building the system that builds the system. 

“Two things are happening in this AI world. First, a lot more engineers are coming into the fold, because companies like Replit or Lovable are making software development approachable for people who haven’t been professionally trained. Second, there’s going to be a lot more applications written than ever before. The problem space is shifting toward how we run all those applications safely and reliably, with all the guardrails. That’s where the industry is headed.”

France’s OVHcloud is moving beyond cloud infrastructure into frontier AI model development, a shift that could test whether Europe can produce another serious alternative to US and Chinese AI systems.

The company, one of Europe’s leading homegrown cloud providers, plans to train a family of models from scratch and aims to open-source them once they meet its performance targets, CEO Octave Klaba told Reuters.

The move would put OVHcloud in closer comparison with Mistral AI, the Paris-based model developer that has become Europe’s most visible challenger to US AI labs.

Klaba said the economics of building advanced AI models have changed, with improvements in chips, training methods, and synthetic data reducing the cost of a project that may once have required about $1.15 billion (€1 billion) to now cost less than $230 million (€200 million).

Reuters reported that OVHcloud said one of its models has completed pre-training on Jupiter, the Germany-based EuroHPC supercomputer described as Europe’s fastest and its first exascale system, though the company has not yet disclosed detailed performance benchmarks.

This comes as European governments and enterprises are increasingly having to assess AI infrastructure through the lens of data governance and continuity of access, rather than performance alone.

Those concerns were sharpened this month after Anthropic said a US government export-control directive required it to suspend access to its Fable 5 and Mythos 5 models by foreign nationals inside and outside the US.

Training is only the opening cost

OVHcloud’s lower cost estimate does not capture the full cost of becoming a frontier AI model provider, said Neil Shah, vice president for research and partner at Counterpoint Research.

The $230 million (€200 million) figure likely refers mainly to the initial training run, Shah said. Once trained, however, models require continued investment because they can become depreciating assets if they are not improved with fresh data.

OVHcloud would also need to spend on fine-tuning, post-training, sovereign infrastructure, storage, security, distribution, and enterprise support. It would also need enough scale to make model serving economically viable against established AI providers such as Google and Anthropic.

“Model is seen as a depreciating asset if it is not consistently trained and kept fresh with the data,” Shah said.

That makes OVHcloud’s plan a test not only of technical capability, but also of policy support and economic viability. If the company falls short, enterprises may be reluctant to shift workloads away from more established models.

The lower training cost could still give OVHcloud a credible starting point, said Charlie Dai, principal analyst at Forrester.

The budget range can be enough to produce a credible frontier model as efficiency gains reduce the cost of entry, Dai said. But enterprise competitiveness will depend on sustained capabilities beyond training, including inference efficiency, data pipelines, evaluation frameworks, and ecosystem reach.

Buyers need proof

OVHcloud’s plan remains an expression of intent rather than demonstrated capability, said Sanchit Vir Gogia, chief analyst at Greyhound Research, pointing to the absence of published benchmarks and other details.

“$200 million now buys a serious training run,” Gogia said. “It does not buy a serious enterprise AI franchise.”

Gogia said questions around sovereignty also extend to the infrastructure used to train the model, noting that pre-training was run on Jupiter rather than on infrastructure owned or controlled by OVHcloud.

The system is a publicly owned European supercomputer in Germany that runs on American silicon, Gogia said, adding that this shows how partial European AI sovereignty remains.

CIOs will need evidence that the models can be supported in production, governed effectively, audited when needed, and exited without major disruption.

Gogia said a European-owned model could reduce some dependence on US and Chinese providers, but would not remove jurisdictional risk. “Sovereignty does not abolish the off switch,” he said. “It changes whose hand rests upon it.”

OVHcloud’s move into model development could also alter the lock-in risks enterprises need to assess, Gogia said. Customers may be able to move cloud infrastructure later, but find it harder to shift AI workloads once applications and processes are built around a provider’s models and governance tools.

The article originally appeared on ComputerWorld.

France’s OVHcloud is moving beyond cloud infrastructure into frontier AI model development, a shift that could test whether Europe can produce another serious alternative to US and Chinese AI systems.

The company, one of Europe’s leading homegrown cloud providers, plans to train a family of models from scratch and aims to open-source them once they meet its performance targets, CEO Octave Klaba told Reuters.

The move would put OVHcloud in closer comparison with Mistral AI, the Paris-based model developer that has become Europe’s most visible challenger to US AI labs.

Klaba said the economics of building advanced AI models have changed, with improvements in chips, training methods, and synthetic data reducing the cost of a project that may once have required about $1.15 billion (€1 billion) to now cost less than $230 million (€200 million).

Reuters reported that OVHcloud said one of its models has completed pre-training on Jupiter, the Germany-based EuroHPC supercomputer described as Europe’s fastest and its first exascale system, though the company has not yet disclosed detailed performance benchmarks.

This comes as European governments and enterprises are increasingly having to assess AI infrastructure through the lens of data governance and continuity of access, rather than performance alone.

Those concerns were sharpened this month after Anthropic said a US government export-control directive required it to suspend access to its Fable 5 and Mythos 5 models by foreign nationals inside and outside the US.

Training is only the opening cost

OVHcloud’s lower cost estimate does not capture the full cost of becoming a frontier AI model provider, said Neil Shah, vice president for research and partner at Counterpoint Research.

The $230 million (€200 million) figure likely refers mainly to the initial training run, Shah said. Once trained, however, models require continued investment because they can become depreciating assets if they are not improved with fresh data.

OVHcloud would also need to spend on fine-tuning, post-training, sovereign infrastructure, storage, security, distribution, and enterprise support. It would also need enough scale to make model serving economically viable against established AI providers such as Google and Anthropic.

“Model is seen as a depreciating asset if it is not consistently trained and kept fresh with the data,” Shah said.

That makes OVHcloud’s plan a test not only of technical capability, but also of policy support and economic viability. If the company falls short, enterprises may be reluctant to shift workloads away from more established models.

The lower training cost could still give OVHcloud a credible starting point, said Charlie Dai, principal analyst at Forrester.

The budget range can be enough to produce a credible frontier model as efficiency gains reduce the cost of entry, Dai said. But enterprise competitiveness will depend on sustained capabilities beyond training, including inference efficiency, data pipelines, evaluation frameworks, and ecosystem reach.

Buyers need proof

OVHcloud’s plan remains an expression of intent rather than demonstrated capability, said Sanchit Vir Gogia, chief analyst at Greyhound Research, pointing to the absence of published benchmarks and other details.

“$200 million now buys a serious training run,” Gogia said. “It does not buy a serious enterprise AI franchise.”

Gogia said questions around sovereignty also extend to the infrastructure used to train the model, noting that pre-training was run on Jupiter rather than on infrastructure owned or controlled by OVHcloud.

The system is a publicly owned European supercomputer in Germany that runs on American silicon, Gogia said, adding that this shows how partial European AI sovereignty remains.

CIOs will need evidence that the models can be supported in production, governed effectively, audited when needed, and exited without major disruption.

Gogia said a European-owned model could reduce some dependence on US and Chinese providers, but would not remove jurisdictional risk. “Sovereignty does not abolish the off switch,” he said. “It changes whose hand rests upon it.”

OVHcloud’s move into model development could also alter the lock-in risks enterprises need to assess, Gogia said. Customers may be able to move cloud infrastructure later, but find it harder to shift AI workloads once applications and processes are built around a provider’s models and governance tools.

In hardware, when you ship something broken, the consequences are severe and often irreversible. That’s the world I worked in for years, in verification roles at Mellanox and later at Alibaba. The stakes forced the industry to build a rigorous verification culture. You proved designs worked before they left the building.

In software, verification disciplines look like CI/CD pipelines, static analysis, canary deployments, and observability. But those systems were built around code written at human speed, with human comprehension baked into the process. AI code generation has broken that assumption. The writing process can no longer be trusted to carry institutional knowledge and judgment into the codebase. The industry is being pushed toward the kind of rigorous verification culture that hardware engineers have practiced for decades.

Enterprises are generating code faster than at any point in history. Google recently disclosed that 75% of the company’s new code is now AI-generated. Meta has set internal targets requiring most of its engineers to generate the majority of their committed code with AI tools by mid-2026. The velocity gains are significant. But a growing body of evidence suggests the industry is accumulating a new form of technical debt, one that is less visible than the traditional kind and harder to unwind. It’s also preventable, and the organizations that get ahead of it will have a meaningful advantage over those that don’t.

Understand what makes this debt different

The standard narrative frames this as AI writing bad code. That’s not quite right. The more precise problem is cognitive debt: the loss of understanding of how and why software was built the way it was.

When a human writes code, something else happens alongside the typing. They simulate edge cases, reason through dependencies, and make judgment calls grounded in organizational context, including the business requirements behind a feature, the best practices the team has established, and the reasoning behind past architectural choices. That cognitive loop is how institutional knowledge gets built. When AI writes the code, you can get output that is syntactically correct, passes CI, ships cleanly, and leaves no one holding the mental model. The code works until something changes or breaks, and then the team is excavating a black box.

This is distinct from traditional technical debt, which is messy code. Cognitive debt is invisible code that functions but that nobody truly owns. And it compounds faster, because the same velocity that makes AI generation attractive is what prevents anyone from stopping to build the understanding that maintainability requires.

GitClear’s analysis of 211 million changed lines of code across major repositories found that during 2024, duplicate code blocks of five or more lines increased eightfold, while refactoring dropped from 25% to under 10% of all code changes. Refactoring is the slow, unglamorous work that keeps codebases healthy, and developers are doing far less of it. Google’s 2024 DORA report found that a 25% increase in AI adoption correlates with a 7.2% decrease in delivery stability. DORA analysts note that the root cause isn’t flawed code per se; AI inflates batch sizes, and larger changesets have always been riskier to ship.

These findings aren’t indictments of AI-assisted development. They’re diagnostics, and they point toward a specific set of fixes.

Close the context gap first

In a survey of 609 developers we conducted last year, 65% said AI misses relevant context during critical tasks like refactoring, writing tests, or reviewing code. Context is the primary driver of AI code quality, and it’s where most enterprise organizations are underinvesting.

When an AI tool generates code without access to your organization’s architectural decisions, historical pull requests, security policies, or existing module patterns, you get solutions that are locally correct but globally incoherent. Closing that gap requires context engineering: ensuring the tools and agents you use have access, at the right moment, to the right organizational knowledge, and the judgment to determine what is actually relevant for a given task. A retrieval system that surfaces too much irrelevant context can degrade output quality as readily as one that surfaces too little. The specific tooling matters less than the discipline. Context infrastructure needs to be actively maintained, not indexed once and forgotten.

Build this infrastructure before you scale AI generation. Retrofitting is significantly harder. Treat it the way you treat your CI pipeline, as a prerequisite for safe production deployment.

Consider what happens when a team has built this context infrastructure well. Their code review tooling knows about a deprecated internal API, because that deprecation decision lives in months of past pull request discussions that have been indexed and surfaced. When generated code references the old API, the review flags it. Without that context layer, the same mistake gets waved through every time. That’s the kind of institutional knowledge that evaporates when humans stop writing every line of code, and that you have to actively work to preserve.

Build a verification layer that matches your generation velocity

Almost all of the investment in AI-assisted development has gone into generation. Very little has gone into verification. That imbalance is where the tech debt accumulates.

I think of these as the blue team and the red team. The blue team covers code generation, autocomplete, and agentic coding. It’s getting the headlines, the budgets, and the product launches. The red team covers integrity checks, behavior coverage, and alignment with organizational standards. In most organizations, it’s an afterthought. A CI pipeline catches obvious failures. A code review might happen, but reviewers are overwhelmed by the volume of AI-generated output and cannot meaningfully evaluate all of it. The result is code with a veneer of having been reviewed without anyone having actually understood it.

The Crowdstrike outage of 2024 is worth keeping in mind here. AI didn’t generate the problematic code, but the incident illustrated what happens when a single software error propagates through production systems without sufficient verification. That exposure multiplies when code is being generated faster than humans can understand it.

A real verification layer means automated analysis that evaluates whether generated code aligns with your organization’s best practices, architectural standards, and compliance requirements. It means test coverage that reflects intended behavior, not only the happy path the AI chose to generate tests for. And it means traceability: a connection between the requirement and the implementation, so that six months from now, someone can understand what the code does and why it exists.

The numbers support investment here. In the same developer survey, teams that integrated AI into their code review workflow saw quality improvements in 81% of cases, compared to 55% for comparable teams without it. 

Make ownership non-negotiable

Every piece of AI-generated code in production needs an accountable human who understands it well enough to maintain it. This is harder than it sounds, and it’s where most organizations are falling short.

The same velocity that makes AI generation attractive also creates pressure to skip the slow work of genuine comprehension. A developer reviews a 500-line pull request that an AI generated in three minutes and faces a real choice: spend two hours actually understanding it, or approve it because it looks right, passes the test, and “LGTM” (looks good to me). 

Real ownership means slowing down generation velocity enough to allow for meaningful review, and being explicit with your team that this is the right trade-off. When that doesn’t happen, you’ve started building your next legacy system.

What to do this quarter

The good news is that none of this requires a multi-year transformation. The structural problems are real, but they have concrete solutions, and engineering leaders can make meaningful progress on all three fronts without waiting for the next budget cycle.

• Audit your context infrastructure. If your AI tools are generating code without access to your organization’s architectural decisions, deprecated APIs, and security policies, fix that before expanding generation velocity further. The quality of context determines the quality of output.
• Invest in the red team. Establish an automated code quality and governance layer that evaluates generated code against your specific organizational standards, beyond functional correctness. This is a distinct investment from your code generation tooling and needs to be treated as such.
• Establish ownership explicitly. Every AI-generated feature in production should have a human owner who genuinely understands it. Make that a formal requirement.

The organizations that get this right will find that AI generation becomes far more reliable once it has a verification layer underneath it. The ones that don’t will keep shipping faster while understanding their systems less, until the accumulated debt forces a reckoning. That’s a solvable problem. The question is whether you solve it now or later.

—

New Tech Forum provides a venue for technology leaders—including vendors and other outside contributors—to explore and discuss emerging enterprise technology in unprecedented depth and breadth. The selection is subjective, based on our pick of the technologies we believe to be important and of greatest interest to InfoWorld readers. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Send all inquiries to doug_dineley@foundryco.com.

Because generative AI (genAI) tools and services have become so ubiquitous (and popular), the costs of using them are going through the roof — leading to an insatiable appetite for tokens.

Tokens represent a common way to measure and price AI use. Much like letters and words in English, large language models (LLMs) grasp a sentence or query by breaking words into tokens.

With the AI explosion well under way, tokens are now “the fundamental units of data our models process, many representing a problem being solved,” according to Google CEO Sundar Pichai. (Google, by the way, processes about 3.2 quadrillion tokens a month.)

But as the price of all those tokens adds up, business and IT execs are looking for ways to cut costs while keeping corporate productivity up. Uncontrolled token use has already landed one company with an unexpected $500 million AI bill. 

There are a number of ways companies can rein in the price of AI at the model, infrastructure, silicon, and business levels. Here’s a look at how some of those savings might actually be achieved.

Switch to lower-cost models

One way of potentially saving money is by re-routing AI work to a cheaper model, Pichai said. At Google that would Gemini 3.5 Flash. It delivers “frontier-level capabilities at less than half the price of comparable frontier models.

“If companies use a mix of [Gemini 3.5] Flash and other frontier models, they could save a lot of money,” Pichai said.

Those kinds of models provide cheaper tokens, with reasoning that’s good enough for many users — if not as strong as mainstream Gemini 3.5 — to deliver useful results.

“There is sometimes overkill with the [LLMs],” said Deepak Seth, senior director analyst at Gartner. “I don’t always need a large language model which has been trained on the works of Charles Dickens and Shakespeare and Harry Potter.”

Hyperframe Research principal analyst Steven Dickens can’t stop using Amazon’s Quick, which costs $20 a month, for personal tasks. “It is great personal ROI as it has not only made tasks faster, but unlocked tasks I would never have even attempted previously,” Dickens said.

Don’t forget the hardware and software part of the equation

The token crisis isn’t new, said Dheeraj Pandey, CEO of DevRev, who likens what’s going on now in the AI market to the disruptions that emerged with the arrival of cloud computing and virtualization years ago. 

“We let chaos reign and then we had to rein in the chaos,” Pandey said. “The word that people started using was server consolidation and virtualization.”

The answer to the token problem, he said, is the same: “Anything in systems can be solved with caching and indirection.”

DevRev, for example, is building a memory layer between AI agents and primary data sources, such as Salesforce or ERP records; that can cut token load and make data movement more efficient. The layer holds a knowledge graph with answers to common agent questions and runs on cheaper CPUs, avoiding more costly GPU cycles.

Sending agents straight at systems like ServiceNow and Salesforce “will burn a lot more tokens. It’s also not precise. And finally, it’s not safe enough where I can roll it back in case an agent has committed a mistake,” Pandey said.

Network automation firm NetBrain uses a different method: It uses conventional computing to map a network’s layout then feeds only key information to models for planning and reasoning, where AI excels. “So you don’t have to spend all the tokens,” said Netbrain CTO Song Pang.

Focus on prompt efficiency

Staffing firm ManpowerGroup has found that prompt efficiency can be an effective tool for improving token use, both internally and externally for clients.

For example, users accessing its internal labor-market tool initially needed 10 follow-up questions to drill into a query. A year later, more efficient use of prompts has brought that number down to an average of four, said Max Leaming, head of data science and AI solutions at ManpowerGroup. 

“They’re using fewer tokens and they’re simply more efficient,” he said. “And that in large part has to do with your ability to prompt efficiently.”

Go local

New AI hardware that generates free tokens at home could ease some of the cost crisis.

At GTC Taipei earlier this month, Nvidia and Microsoft unveiled RTX Spark, an agentic AI desktop PC that runs agents and 120-billion-parameter models locally on Windows. The goal is “to deliver unmetered intelligence to every home and every desk with Windows,” Microsoft CEO Satya Nadella said in a statement.

Some companies are looking to reduce cloud AI costs by putting their own hardware in data centers, with vendors such as HPE and Dell providing servers installed in independent facilities. (On-premise AI is gaining ground amid sovereign AI and geopolitical concerns, including the recent conflict in the Middle East, where large data centers were struck with missiles.)

“There are local, region-specific and multiple vendor AI solutions. All of those things can help mitigate the risk. But they’re not going to eliminate it,” said Max Goss, senior director analyst at Gartner.

Use forward-deployed engineers

Reducing token costs is something that may fall to forward-deployed engineers (FDEs) in customer environments, said Taimur Rashid, managing director of AWS’s Generative AI Innovation Center.

“I expect these teams to be able to architect systems that have those cost requirements in mind, whether it’s use a different model or a different use case that doesn’t increase the per-token cost,” Rashid said.

Companies may spend heavily on token consumption, “but if you’re generating revenue, as long as the economics work out, then you’re at peace,” Rashid said.

The use of FDEs is gaining ground as IT decision-makers look to both rollout successful AI deployments while also keeping an eye on costs.

Change the measure of success from tokens to outcomes

Even with the current emphasis on reducing token use to save money, the metrics used to measure AI success are likely to shift, Gartner’s Seth said. At some point, token-based pricing will move more toward an outcome-based model, where the unit of value is outcomes, not fragments of words.

“Some companies are moving towards outcome-based pricing,” Seth said. “When people start realizing the real cost of tokens, then companies will start looking at token efficiency.”

Adobe’s Firefly Graph is now available to Creative Cloud customers, offering a node-based workflow tool designed to help business create content at scale with generative AI (genAI). 

With Firefly Graph, users can connect multiple tools in visual workflow, with each “node” performing a specific task before passing its output to the next node. This gives creative professionals more control over generated outputs, according to Adobe, and makes it easier to try out ideas by swapping, adjusting or adding components.

For example, a user could start with a text prompt box that connects to a node that generates an image using an AI model from Adobe or third-parties such as Google and OpenAI. Further along the chain, the user could add nodes to remove a background or upscale an image, for instance, before producing an image, video or other asset ready for use.

Changing one aspect, such as adding a reference image or adapting the text prompt, would change the final output.

It’s an approach similar to node-based workflow tools such as ComfyUI — a startup valued at $500 million which claims more than 4 million users. Others include Weavy, acquired by Figma last year for a reported $200 million. 

With so many AI tools available to creative professionals, workflows can get complex and hard to replicate, said Elliot Sedegah, director for strategy and product marketing at Adobe. Firefly Graph provides access to more than 300 different node types, including images, video editing and AI generation tools across Adobe’s portfolio and third-party tools. 

“Whether you’re working at a mom-and-pop shop or a larger enterprise, you’re looking for consistency and then bringing that into a workflow so that you’re not hopping in and out of different tools,” he said. “Putting all that together takes massive amount of time, and sometimes it’s very difficult to even know what you did.”

Once created, workflows can be shared across an organization as repeatable processes for other individuals or teams to use. “Think of that rock star creative that you have and the recipes they create: those are now canonized as workflows, as assets, that the rest of the organization can take and reuse over and over again,” said Sedegah.

In addition, while creative professionals are needed to created high quality assets, reusable workflows can be put into the hands of broader teams to create content for large audiences, said Sedegah.

Firefly Graph addresses a challenge that most large creative organizations face, said Lisa Gately, principal analyst at Forrester — namely that their best creative workflows “live inside the heads of a few experts.

“Teams can generate images and video with AI, but reproducing the exact sequences of creative decisions, model selections, edits, and refinements that lead to a high-quality result is difficult and inconsistent. Firefly Graph turns those workflows into reusable assets,” she said.

While other node-based workflows aim to address similar problems, Adobe’s pitch is that Firefly Graph provides customers with the benefit of integration into its product suite. 

“Firefly is a full, broader AI creative studio, not just a node-based tool, so [Firefly Graph] is a part of a bigger picture,” said Sedegah. “The strength is having everything in one place with the tools that people know.” 

“Where Adobe differentiates is in enterprise integration,” said Gately, with Adobe connecting Firefly Graph to a range of other Adobe tools. Those include Creative Cloud applications; Firefly Boards for ideation; and Firefly Creative Production. 

“The workflow becomes part of a broader content supply chain instead of a standalone creation tool,” she said. ”Organizations committed to other tools are unlikely to migrate for a node-based canvas — making a change is about the broader content supply chain.”

Firefly Graph is available now to Adobe Creative Cloud for Enterprise subscribers. Creative Cloud for Enterprise is licensed by seat, requiring custom, volume-based enterprise agreements. Creative Cloud for Enterprise customers receive credits for Firefly Graph. Creative Cloud for teams customers can sign up for a public beta online. Creative Cloud pricing starts at $99.99 per license each month.

This article was updated on June 19 with pricing information from Adobe.

Microsoft has introduced usage-based billing for Copilot Cowork, which is now generally available. 

Microsoft unveiled Copilot Cowork in March, pitching it as an AI agent that’s capable of independently performing long-running, multi-step tasks — even when a user’s computer is off. 

It’s built on the same technology that underpins Anthropic’s Claude Cowork. Unlike Claude Cowork, which can interact directly with files and applications on a user’s computer, Copilot Cowork runs in Microsoft’s cloud environment and acts on documents held in a customer’s Microsoft 365 tenant.

On Tuesday, Microsoft unveiled pricing details for Copilot Cowork, which involves usage-based billing in addition to a Microsoft 365 Copilot license ($30 per user each month for large enterprises before discounts, and $20 for Microsoft 365 Copilot for Business). 

The usage-based pricing is calculated from four components, according to Microsoft: “model use, context retrieval, tool calls, and runtime.” 

In practice, this means more intensive tasks that draw on multiple sources, use “deep reasoning” and generate two or more outputs will lead to higher costs — denoted in Copilot Credits. 

There are two payment options: pay as you go — priced at 1 cent per credit — and P3, where customers commit to usage volume in advance and receive a discount.

Cowork is turned off by default; IT admins can decide when to make it available and which employees get access. Admins can also impose spending limits at the tenant, group, and user level, and receive notifications when spending reaches a certain level.

Users will also be able to see how much each tasks costs in credits (available “soon after” the general availability launch, Microsoft said). 

Copilot Cowork customers can select from multiple AI models, including Anthropic’s Opus 4.8 and Sonnet 4.6, while those enrolled on the Frontier program can access OpenAI’s GPT 5.5 and Microsoft’s own Cowork 1 model.

Microsoft is considering hosting a version of DeepSeek’s open source models. “We are actively exploring a range of options that meet those requirements, and DeepSeek is one of several models under consideration. We’ll share more specifics on the underlying model closer to release,” said a Microsoft spokesperson.

Microsoft also announced new integrations with third-party apps such as Miro and Monday.com, with more — such as Adobe, Box, and Canva — coming soon.

This article was updated on June 19 to add Microsoft’s comment on DeepSeek.