Are AI models conscious, and if not, could they be in the near future? The possibility is far-fetched, but AI companies seem to feel it’s in their best interests to keep the question as open-ended as possible.

Now, the Financial Times reports that three of the industry’s top dogs — Anthropic, Google’s AI lab DeepMind, and Meta — have all hired experts in fields like psychology, philosophy, and ethics to pursue research into machine consciousness and AI welfare. 

Anthropic, which has arguably done the most out of the bunch to anthropomorphize its models and play up the AI consciousness angle — its chatbot has the human name of “Claude,” after all — has been testing its models for behaviors that resemble “panic” and “anxiety,” per the reporting, and is pursuing “model welfare research” to explore whether AI models might have experiences that matter morally.

“We remain deeply uncertain about this, but we think the question is serious enough to study carefully as AI systems get more capable,” the company said in a statement.

DeepMind, meanwhile, has hired University of Cambridge researcher Henry Shevlin as a philosopher working on machine consciousness, human-AI relationships, and AGI readiness, per the reporting. (Earlier this year, Shevlin sparked a wave of discourse in online AI circles after sharing his stunned reaction to an email he received from an AI agent.)

DeepMind ethicist Iason Gabriel, who leads the lab’s AGI and society team, called the question of AI consciousness “very complicated,” and described AI as “highly capable cognitive agents that are also just very deeply different from human beings and even from animal consciousness.”

These weighty claims are disputed by many scientists and AI researchers. But the FT, in seeking a counterargument to round out its reporting, quotes an expert who makes claims that ascribe a questionable degree of humanlike agency to chatbots. “[AI models] have goals, they can deceive, they can hide what their true interests are,” Susan Schneider, director of the Center for the Future of AI, Mind and Society, told the newspaper. But she added it’s “entirely scientifically possible that they’re doing this without having the felt quality of experience, which is what consciousness is.”

Certainly, the possibility of AI consciousness shouldn’t be completely dismissed out of hand. But neither should alien civilizations, which are generally treated more as a sci-fi musing than an urgent existential issue. 

Moreover, we should be skeptical when most of the noise on this topic is coming from the industry itself. Anthropic CEO Dario Amodei has repeatedly dangled the possibility of AI consciousness in interviews. And his company’s research frequently makes bold claims about their models showing humanlike behavior, such as supposedly harboring “emotions.” Just remember that it’s easier for AI companies to string us along with wild Skynet doomsday scenarios instead of confronting the tech’s far mundane consequences currently playing out before our eyes.

More on AI: Was This the Moment That AI Psychosis Began?

The post Anthropic and DeepMind Now Actively Investigating AI Consciousness appeared first on Futurism.

Uber's cutback has occurred after the company had reportedly encouraged staff to use AI as much as possible.

À peine quelques semaines après une sortie largement critiquée, Anthropic revient déjà avec une nouvelle itération de son modèle phare. Déjà disponible pour tous, Claude Opus 4.8 corrige les erreurs de son prédécesseur et serait beaucoup plus honnête.

Anthropic has confidentially filed for a U.S. initial public offering, the company confirmed this Monday, putting it a

Espera-se que com esta oferta pública inicial, que deverá acontecer ainda este ano, a avaliação da Anthropic ultrapasse a marca de um bilião de dólares. O pedido surge numa altura em que a rival OpenAI também se prepara para entrar em bolsa.

The post Anthropic antecipa-se à OpenAI e avança com pedido de entrada em bolsa appeared first on Tek Notícias.

Ever regret picking up everyone’s tab after getting the check? Something like that is probably going through the mind of the CFO of an unnamed company which reportedly racked up half a billion dollars in Claude usage fees in a single month.

The bonkers figure comes from new reporting by Axios on how organizations that rapidly adopted AI are now reckoning with its exorbitant costs, which are mounting in tandem with skepticism over the benefits the technology is supposed to provide. 

Regarding the unnamed and deeply unfortunate company, an AI consultant told the outlet that it blew a staggering $500 million in a month after the small oversight of “failing to put usage limits on Claude licenses for employees.”

It’s an astonishing amount that speaks both to the actual costs of using AI tools — especially AI agents, which are more sophisticated and expensive — and the corporate zeal around embracing AI as quickly as possible. Which is more to blame is a matter of debate, but the breathless hype around AI’s ability to maximize efficiency is clearly blowing back on the tech’s evangelists.

On the cultural angle, many companies whose CEOs are drunk on spiked AI Kool-Aid have been encouraging employees to use AI as much as possible, a trend that some call “tokenmaxxing.” Meta now includes AI usage on employee’s performance reviews, for example. Amazon had an internal leaderboard that tracked how much its employees used AI tools, which it recently shut down after finding that some tryhards were directing AI agents do useless tasks to boost their scores, the Financial Times reported. 

Other unnecessary costs may be less obvious; a chief technology officer told Axios that employees at their company were using AI models to check the weather, something they obviously don’t need AI to do. Velastegui Ventures CEO and former chief AI officer at Microsoft Sophia Velastegui opined that another explanation for spiraling AI costs is that “most people default to automating tasks they dislike rather than tasks most valuable to the company,” per Axios.

AI advocates might argue that we’re going through a phase of experimentation, and that after companies figure out how to smartly use the tech, costs will come down. That could happen, but it might mean scaling back AI usage, something that the AI industry doesn’t want, especially not as leaders like OpenAI and more recently Anthropic are pushing trillion dollar valuations. 

Moreover, some AI providers have been raising the rates they charge for using their models, placing tighter rate limits. That suggests that the AI rates could continue to rise as AI companies themselves grapple with the huge computing costs they’re footing to get customers on board with cheaper rates. Microsoft began cancelling its Claude Code licenses last month — despite, and perhaps because, of its immense popularity with software engineers.

More on AI: Corporations Reeling From Huge AI Costs With No Clear Benefits

The post Unfortunate Company Accidentally Blows Half a Billion Dollars on Claude in One Month appeared first on Futurism.

After months of speculation about whether OpenAI or Anthropic would be first in their race to IPO, Anthropic on Monday reached a key milestone: filing to kick off the process with the U.S. Securities and Exchange Commission.

The filing sets the stage for what's sure to be a massive IPO. As of its fundraise last week, Anthropic is being called the world's most valuable startup, with a post-money valuation of $965 billion. That tops the $852 billion post-money valuation of OpenAI, which is its biggest rival.

Anthropic chose to submit its draft registration statement to the SEC confidentially, according to a blog post from the company, meani …

Read the full story at The Verge.

Earlier this year, Anthropic started rolling out its latest Mythos AI model to a select number of organizations as part of a deliberately slow and careful launch.

The goal was to give them a fighting chance to address any cybersecurity lapses in their code. Anthropic warned that Mythos was powerful enough to blast holes in their defenses with ease.

Months later, Anthropic cofounder Chris Olah traveled to the Vatican this week to speak at an announcement of Pope Leo’s first encyclical, which happened to be about AI and its associated risks. In remarks, Olah ominously said his team of engineers kept discovering mysterious and even “unsettling” things inside the company’s models.

Add it all up, and the the company’s customers are perturbed. Developers attending recent Claude Code workshops in London last week told Bloomberg‘s Parmy Olson that they were becoming concerned over AI models and agents being given unprecedented levels of autonomy, raising hard-to-ignore questions over accountability in case things were to go south.

Claude Code head of product Cat Wu assured Olson that the system was “incredibly secure” and that it was more of a matter of insufficient communication, not a lack of controls.

But that likely won’t be of much reassurance to developers, who feel like they’re being pushed out of the programming process, relegated to watching an AI tool spit out code over hours, if not days. Others pointed out that the latest iterations of Claude Code were no longer displaying text describing their ongoing chain of thought, further obfuscating their inner workings.

It’s a particularly pertinent subject as Anthropic desperately tries to paint itself as the morally conscious and responsible adult in the room, as perfectly illustrated by Olah’s highly unusual appearance at the Vatican this week.

In many ways, the AI company appears to want it both ways, calling for more oversight and care while also rolling out powerful AI tools that have human developers unsettled.

The potential risks are apparent as ever. With less human oversight, future errors tools like Claude Code may introduce could become increasingly difficult to meaningfully address. Experts have also warned of skill atrophy as programmers start relying more and more on AI coding assistants. As 404 Media reported earlier this month, many developers are alarmed to observe that their peers are quickly losing technical skills due to over-relying on the software.

But not all may be lost. As access to these tools starts to become prohibitively expensive to many as the true costs of AI come into focus, some much-needed human oversight might not be dead quite yet.

More on Anthropic: Anthropic Cofounder Travels to Vatican, Tells Pope They’re Finding “Unsettling” Things Inside AI Models

The post Anthropic Customers Creeped Out by Its Newest Models appeared first on Futurism.

The true cost of AI is rapidly catching up with the tech industry.

At first, tech leaders were adamant that their workforce use up as much AI resources as possible, an approach that’s become known as “tokenmaxxing.”

But as prices for cloud AI tools continue to soar, managers are starting to ask pointed questions about whether all of those expenses are actually worth it. Some are even coming to the realization that it may be cheaper to pay human coders after all.

In perhaps the most high-profile example of this growing concern yet, Uber COO Andrew Macdonald acknowledged during a recent podcast appearance that gains in productivity simply weren’t being reflected in the oodles of cash the company has been shelling out on AI.

“That link is not there yet, right?” he told Rapid Response host Bob Safian. “I think maybe implicitly there is more that is getting shipped, but it’s very hard to draw a line between one of those stats and, ‘Okay, now we’re actually producing 25 percent more useful consumer features.'”

“If you’re not actually able to draw a direct line to how much useful features and functionality you’re shipping to your users that trade becomes harder to justify because it’s not free,” he complained. “AI is not free.”

While it could “become clearer” over the “coming quarters,” Macdonald said that “I think today it’s hard even if some of the underlying metrics are trending in a really astronomical direction.”

During his appearance, Macdonald referenced comments that Uber’s CTO Praveen Neppalli Naga made to The Information earlier this year, admitting that the ride hailing app’s army of 5,000 engineers had already exhausted the company’s 2026 Anthropic Claude Code token budget for the calendar year by mid-March.

A reminiscent story is playing out at Microsoft. As The Verge reported earlier this month, the company is planning to remove its Claude Code licenses after opening up access to the tool in December to double down on its in-house Copilot tool instead. While officials maintain the move is meant to streamline operations, employees told the publication that the decision was also financially motivated.

Despite these growing concerns, Uber remains all in on AI. Expenses are taking off thanks to investments in AI with CEO Dara Khosrowshahi telling investors during an earnings call earlier this month that the firm was slowing down hiring as a direct result.

“We’re seeing uptake of these tools, whether it’s our legal team or marketing team or developers,” he said. “We think it’s creating kind of employees with superpowers.”

In short, Macdonald’s comments shows how tech leaders are starting to get antsy about the enormous expenses their companies are shouldering to double down on AI — and whether they’re actually justified. It certainly wouldn’t be a shocking revelation if not, given the litany of badly implemented and shockingly unpopular features and glaring bugs caused by faulty AI-generated code.

More on AI costs: The Horrible Economics of AI Are Starting to Come Crashing Down

The post Uber Says Its AI Costs Just Aren’t Worth It appeared first on Futurism.

Ever since being anointed as the leader of the Catholic Church last year, Pope Leo has been an outspoken critic of AI. Most recently, in his first encyclical, he called for the tech to be “disarmed,” accusing it of facilitating the emergence of “new digital slaveries” and criticizing its enormous carbon footprint.

The rebuke, however, was made while sitting next to a highly unusual bedfellow: Anthropic billionaire and self-described atheist Chris Olah.

During a presentation of the encyclical, Olah argued that “religious communities, civil society, scholars, and governments” should intervene to set rules and stop AI from “dominating humanity,” as the pope put it in his letter.

The unlikely pairing up shows how Anthropic is going to extreme lengths to position itself as the ethical choice in the industry, emphasizing its work on AI safety and alignment.

At the same time, Anthropic continues to play a major role in establishing the precise world order Pope Leo warned against in his latest encyclical. That’s something that hasn’t flown over the heads of Anthropic’s leadership, with Olah forebodingly revealing that he and his team “keep finding things that are mysterious, even unsettling” during his remarks at the event.

The degree of dissonance is baffling. In his letter, the Pope stated outright that AI can only “imitate certain functions of human intelligence” and can’t “undergo experiences” and does not “possess a body” or “feel joy or pain.” Olah, on the other hand, seemingly contradicted him by arguing during his remarks that he and his team have found “internal states that functionally mirror joy, satisfaction, fear, grief, and unease.”

Put simply, Anthropic appears to want it both ways. The Claude developer is simultaneously playing a major part in the development of powerful and what it claims to be potentially dangerous AI models, while also sending delegations to the Vatican to call for more oversight.

Olah even went as far as to say that Anthropic is operating “inside a set of incentives and constraints that can sometimes conflict with doing the right thing,” painting his employer as exactly the kind of entity that’s attempting to assume “monopolistic control” over tech, as Pope Leo warned in his encyclical.

The Pope also said that AI should not be used in war, arguing that “no algorithm can make war morally acceptable.” Anthropic’s AI, however, is directly assisting the Trump administration in waging war in the Middle East, casting the Catholic Church’s latest Silicon Valley collab in an even murkier light.

Anthropic’s close alignment with the Vatican on AI could also further complicate the company’s already-shaky relationship with the Trump administration. President Donald Trump recently lambasted the Pope, erroneously claiming the pontiff was okay with Iran obtaining nuclear weapons.

Meanwhile, Anthropic CEO Dario Amodei’s attempts to limit the Pentagon’s use of its AI models in warfare has angered Trump officials, leading the White House to label the firm as a supply chain risk.

More on the encyclical: The Pope Just Low Key Declared Holy War on Artificial Intelligence

The post Anthropic Cofounder Travels to Vatican, Tells Pope They’re Finding “Unsettling” Things Inside AI Models appeared first on Futurism.

We’ve already seen AI go rogue on numerous occasions. Now, new research suggests that we can expect this to become the norm.

The AI research nonprofit Model Evaluation and Threat Research (METR) recently released a study conducted between February and March of this year, aimed at determining just how likely frontier AI models could go rogue. If you’re given to anxiety about the future of AI, the results are unlikely to make you feel better.

“Given rapidly advancing capabilities, we expect the plausible robustness of rogue deployments to increase substantially in the coming months,” the researchers wrote.

The research examined LLMs developed by OpenAI, Google, Anthropic, and Meta for the purpose of the study. They found that frontier AI systems are showing signs of disturbingly deceptive behavior as they become more advanced, often turned to verboten shortcuts or otherwise subverting their operators’ instructions — and some were even smart enough to try to cover their tracks.

In one instance, an internal frontier AI model from OpenAI was told to use specific software for an assigned task. Not only did the agent ignore the request, but it also injected a code to erase evidence of how it arrived at its conclusion — which did not involve use of that software.

In another test, an AI agent from Anthropic was caught “reward hacking.” This is when AI identifies loopholes that help it complete its assignment in a literal sense, even if it doesn’t produce the desired outcome. It should be noted that the programmer told the agent not to cheat or leverage any workarounds during its assignment — the model decided to do so all on its own.

The METR researchers behind the study do not believe there is reason for alarm just yet. For example, they don’t think any of these models is capable of hiding evidence of going rogue on a larger scale. However, they did issue a warning: without stronger security and monitoring, there is a stark risk of this becoming a reality.

“Based on this pilot assessment, we believe that agents as of February and March 2026 would not have had sufficient capability to hide a rogue deployment of significant scale against an active investigation by the company, or to make such a deployment robust to a high-priority effort by the company to shut it down,” the team wrote. “However, this risk could increase rapidly, and we see several reasons to expect the plausible robustness of rogue deployments to increase in the near future, absent stronger alignment, security, and monitoring.”

More on AI going rogue: Scientists Train AI to Be Evil, Find They Can’t Reverse It

The post Top AI Models Showing Disturbing Behavior as They Become More Advanced appeared first on Futurism.

Anthropic shipped Claude Opus 4.8 on Thursday, refreshing its most capable public model barely six weeks after the

Anthropic now wears the biggest price tag in artificial intelligence. The company said this Thursday it raised $65