Anthropic has opened the doors to its most advanced AI system yet, but not without placing guardrails around its most sensitive capabilities.

The company on Tuesday unveiled Claude Fable 5, the first publicly accessible version of its Mythos-class AI model. Anthropic says the model outperforms every generally available system it has released to date, delivering major gains in software development, analytical work, visual understanding, and scientific reasoning.

The launch marks a significant shift for the AI firm. Until now, Mythos-level capabilities remained restricted to a small group of trusted organizations because of concerns that the technology could aid cyberattacks and other harmful activities. Fable 5 changes that equation by bringing those capabilities to a broader audience while routing certain high-risk requests through a less permissive model.

Safety before scale

Anthropic said Fable 5 includes a new layer of AI-powered classifiers designed to detect potentially dangerous prompts and jailbreak attempts. The company stress-tested those safeguards before releasing the model to the public.

If users submit requests involving cybersecurity, biology, chemistry, or model distillation, the system automatically redirects those interactions to Claude Opus 4.8 instead of Fable 5. Users receive a notification whenever that fallback occurs, allowing them to understand why the switch happened.

Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.

Its capabilities exceed those of any model we’ve ever made generally available. pic.twitter.com/2AvmEjHIX8

— Claude (@claudeai) June 9, 2026

The company acknowledged that the safeguards may occasionally block harmless requests. However, Anthropic said early testing showed that fewer than 5% of sessions triggered a fallback response. “We’ve deliberately tuned the safeguards to be cautious,” the company said, adding that it plans to refine the system over time to reduce unnecessary interventions.

Built for demanding work

Anthropic argues that Fable 5 delivers its biggest advantages when tackling lengthy and complex assignments. The company says the model can stay focused across extended workflows while maintaining consistency over large amounts of context.

During early evaluations, payments company Stripe reported that the model completed a large-scale migration across a 50-million-line Ruby codebase in roughly one day. According to Anthropic, the same task would have otherwise occupied an engineering team for more than two months.

The company also highlighted gains in financial analysis. On Hebbia’s finance benchmark, Fable 5 posted the strongest results among leading models. Trading firm IMC found that the system performed well across factual analysis, reasoning tasks, root-cause investigations, and expected-value calculations.

Fable 5 also advanced Anthropic’s vision capabilities. The model can extract detailed information from scientific figures and reconstruct web applications from screenshots. In another demonstration, it completed Pokémon FireRed using a minimal vision-only setup after earlier Claude models struggled with the challenge.

Mythos stays restricted

Alongside Fable 5, Anthropic introduced Claude Mythos 5, a version of the same underlying model with fewer restrictions in select areas. Anthropic says the model offers the strongest cybersecurity capabilities it has developed so far.

Access to Mythos 5 remains limited. The company is deploying it through Project Glasswing while expanding availability to approved cyber defenders and organizations responsible for critical infrastructure. Anthropic said it intends to broaden access further through a trusted program in the future.

Fable 5 is available through Anthropic’s Claude API and consumption-based Enterprise offerings. Through June 22, Pro, Max, Team, and seat-based Enterprise subscribers can access the model at no additional cost. After that date, Anthropic plans to move users to a credit-based system before eventually restoring broader subscription access.

As competition intensifies across the AI industry, Anthropic is betting that powerful models do not have to arrive without safeguards. With Fable 5, the company is testing whether frontier capabilities and tighter controls can coexist in the hands of everyday users.

💾

Humanoid robots struggling with tasks like grasping a cup have a new teacher—a person wearing an ultrasound wristband that captures the movement of muscles, tendons and ligaments beneath the skin.

Scientists from ETH Zurich in Switzerland have announced a series of successful experiments involving the repair of injured spinal cords in zebrafish and mice after being injected with magnetically controlled microrobots containing pluripotent stem cells.

According to a statement announcing the ETH team’s work, both the zebrafish and mice injected with the microrobots showed improved movement and normalized exploratory behavior within just a few days after treatment.

Although the initial studies were confined to lab animals, the team hopes their magnetically controlled microrobot approach to the precision delivery of pluripotent stem cells could eventually be adapted for humans with spinal cord injuries.

Spinal Cord Repair Microrobots Treat Nerve Cells That Resist Regeneration

Unlike some cells in the human body, nerve cells within the spinal cord rarely regenerate on their own. If the spinal cord is injured, scarring can further reduce the ability of the nerve fibers within the cord to regrow. For people living with spinal cord injuries, this scenario can offer a bleak prognosis for recovery.

In recent decades, scientists have turned to the natural regenerative properties of pluripotent stem cells. Although the use of stem cells was initially controversial due to the need to extract neonatal stem cells, scientists ultimately learned to convert ordinary stem cells into pluripotent stem cells capable of becoming any cell type in the human body.

The most promising approach involves implanting stem cells at the site of the injury and then using electrical stimulation to promote the growth of new cells. However, according to the ETH team behind the new approach, the current method has several drawbacks.

First, stimulating the cells requires the use of implanted electrodes. Second, the team notes that after transplantation, many of the stem cells “do not always survive or integrate properly into the existing tissue.”

To improve healing and increase the chance of restoring spinal cord nerve cell function, they said their new approach combines therapeutic stem cells with magnetoelectric nanoparticles and uses magnetic fields to guide them to the “precise site of an injury” and “stimulate the stem cells to accelerate repair.”

NPCbots Combine Stem Cells and Magnetically Controlled Nanoparticles

To construct their magnetically guided spinal cord repairing robot, the team built what they termed a “biohybrid microrobot.”

First, the team creates pluripotent stem cells from adult cells by genetically manipulating them in the lab. The result is what researchers call an “induced” pluripotent stem cell (iPS). According to the study authors, these newly pluripotent iPS cells are used to derive living neural progenitor cells (NPCs), meaning “they have the potential to differentiate into various types of nervous system cells.”

Next, the team creates customized magnetic nanoparticles consisting of two layers. The inner layer responds to the “commands” from the magnetic fields used to control them, and an outer layer that converts this response into an electrical signal.

According to Professor Salvador Pané i Vidal of the Multi-Scale Robotics Lab at ETH Zurich, each tiny robot has a reservoir near the center “where we trap the cells.” The professor said the best way to accomplish this is to inject the nanoparticles into a solution with the NPC, “and wait for the two components to bind.”

After 30 minutes, the process is complete, and the team is left with a batch of nanobots, each around 6 micrometers in size. Since the new robots combine a magnetic bot and a reservoir of NPCs, the team named them ‘NPCbots.’

In preparatory cellular experiments, the researchers used hundreds of thousands of NPCbots. However, several million were required for animal trials.

Zebrafish and Mice Regain Movement After Injections

To test the efficacy of their new NPCbots, the ETH team injected millions of them into the spinal cord of zebrafish larvae with spinal cord injuries. To improve the chances of nerve regeneration, the NPCbots were injected directly into the site of the injury.

After injecting the microrobots, the injury site was subjected to electromagnetic fields. Although the repair was not immediately evident, the team noted that after three days, the zebrafish with the injured spinal cords “exhibited nearly normal swimming and exploratory behaviour.”

Next, the team injected millions of NPCbots into mice with completely separated spinal cords. According to the team, the results were “very promising.”

“After 28 days, the animals’ nerve cells had reconnected at the site of the injury,” they explained.

The team said further observation of the treated mice revealed “increasingly normal movement patterns,” including improvements in gait, stride length, and coordination. The ETH team also noted that the mice’s exploratory behavior “improved significantly.”

When discussing the mouse experiments, the team described the result as “particularly significant” because zebrafish spinal cords regenerate, whereas mouse spinal cords do not. They also noted that the treatment was “well tolerated by the animals,” with no reported immune reactions or adverse effects.

Minimally Invasive and More Precise Treatments with Microbots

Unlike science fiction movies, where tiny robots repair tissue mechanically, the research team said that the stem cells carried by the NPCbots are exposed to electrical stimulation, “greatly enhancing their differentiation after transplantation.” This means the nanoparticles convert the electric signals from the EM field into electrical impulses, stimulating healing.

Unlike more complicated stem cell procedures that use implanted electrodes and cables, the team said they only needed to apply external magnetic fields around the spinal cord injury site. The team said this simplification was “crucial,” since the spinal cord is considered extremely sensitive. They also note that the ability to manipulate the frequency and field strength of magnetic fields, combined with their ability to penetrate deeply into tissue, makes the entire process safe and easy to control.

“Microrobotic guidance makes the treatment more precise and minimally invasive,” Hao explained.

After the NPC bots have delivered and stimulated the stem cells, the team said they “essentially dissolve” within the tissue. Because the bots contain a barium titanate coating, the researchers also believe that they will be “stable and minimally reactive” with surrounding tissue. Still, they concede that “further studies will determine whether and how the particles are degraded or excreted over the long term.”

Expanding NPCbot Therapies to Increase the Effectiveness of Other Targeted Therapies

Although the team described the animal trials of their NPCbots as “promising,” they noted that further research is needed before the process can be attempted on human patients with spinal cord injuries. For example, Hao noted that in addition to ‘clinical aspects’ that require more exploration, the team needs to start by determining which magnetic fields “work best in humans” and determining the “optimal stimulation duration” for human NPCs. Nevertheless, the team noted, they are “already considering further applications.”

“The reproducible and scalable production of microrobots using our lab-on-a-chip system demonstrates that the platform’s application potential extends beyond basic research,” Professor Vidal explained.

Some potential biomedical applications mentioned by the RETH team include cardiology, oncology, wound healing, “and other targeted therapies,” adding that their process could make such treatments “safer, more controllable and more effective.”

The study “Magnetoelectric Microrobots for Spinal Cord Injury Regeneration” was published in Nature Materials. 

Christopher Plain is a Science Fiction and Fantasy novelist and Head Science Writer at The Debrief. Follow and connect with him on X, learn about his books at plainfiction.com, or email him directly at christopher@thedebrief.org.

Researchers are replacing rigid silicon-based AI hardware with stretchable, neuromorphic electronics that mimic how the brain processes information, opening new possibilities for long-term human-machine integration. Modern artificial intelligence can outperform humans in tasks ranging from image recognition to medical data analysis, but there is one environment where today’s hardware still struggles: the human body. The [...]

We asked three farmers to tell us how new technology is revolutionizing the way they work.

Os cães-robot podem ser a próxima aposta das forças armadas em Taiwan para reforçar a vigilância e a defesa do território, juntando-se aos esforços de modernização militar face às crescentes tensões com Pequim.

The post Cães-robot podem entrar nos planos de Taiwan para missões de vigilância e defesa militar appeared first on Tek Notícias.

Human hands exhibit unparalleled complexity and dexterity, revealing the mastery of their Creator. Despite advancements in robotics, replicating the nuanced movements and strength of human hands remains a significant challenge. Companies like Proception are striving to develop robotic hands, yet authentic human hand functionality remains unmatched, highlighting the intricacies of divine design.

The world’s most valuable company is chasing Intel and Apple as it tries to bring A.I. agents to laptops and desktops.

A sea-urchin-like robot could offer a new blueprint for making more versatile robots, research suggests.

The contracts, announced on Tuesday, are part of the space agency’s efforts to establish a moon base.