Anthropic Claude Fable 5: Anthropic's new Claude Fable 5 AI model is drawing attention after refusing to answer basic biology questions. The company says strict safeguards are necessary due to its advanced capabilities.
Claude Fable 5 is Anthropic's latest large language model and the first publicly available version of its Mythos-class technology. (Photo: Social Media)
Anthropic Claude Fable 5: Artificial intelligence company Anthropic has introduced Claude Fable 5, a new AI system that the company describes as the first public release from its powerful Mythos class of models. While the launch generated excitement across the technology industry, early users quickly noticed something unusual: the model refused to answer even simple biology-related questions.
The unexpected behaviour has sparked a broader discussion about AI safety, model capabilities and whether advanced safeguards may sometimes prevent legitimate educational use.
Claude Fable 5 is Anthropic's latest large language model and the first publicly available version of its Mythos-class technology.
According to the company, the model possesses highly advanced reasoning and problem-solving capabilities, particularly in technical fields. Anthropic believes the system is powerful enough to perform complex scientific and cybersecurity-related tasks that previous AI models could not easily accomplish.
Because of those capabilities, the company introduced additional restrictions before making the model available to users.
Shortly after the launch, users began sharing examples of unusual responses from the model. In one widely discussed case, a user asked a straightforward biology question about mitochondria.
Instead of providing an answer, the system displayed a warning message, “Fable 5 has safety measures that flag messages on most cybersecurity or biology topics. They may flag safe, normal content as well. These measures let us bring you Mythos-level capability in other areas sooner, and we're working to refine them.”
The message then suggested that users switch to Opus 4.8 for biology-related questions. Reports indicated that similar responses appeared when users asked questions about cancer and other biological topics.
Anthropic says the restrictions are intentional and designed to prevent misuse.
The company argues that highly capable AI systems could potentially assist individuals seeking information related to dangerous biological research.
An Anthropic spokesperson explained, "With the launch of Claude Fable 5, our first Mythos-class model, we believe models now have a greater ability to accomplish real-world scientific tasks and for malicious actors to potentially use our models for highly risky biological research."
The company said it adopted a highly cautious approach while introducing the new model.
Anthropic acknowledged that some harmless educational questions may be blocked under the current system.
The spokesperson added, “We have always used classifiers to block our models from helping with bioweapons-related requests. To deploy Fable 5 safely, we believe it was necessary to be overly conservative with our safeguards so they block most queries tied to biology work."
The company said it plans to improve the system over time to reduce unnecessary restrictions while maintaining safety protections.
Biology is not the only category affected by the new safeguards. Anthropic confirmed that questions related to chemistry, cybersecurity and attempts to analyse or replicate Fable 5's capabilities may also activate safety filters.
When such topics are detected, the system can either refuse to answer or automatically redirect users to another model, depending on the settings selected by the user.
This approach allows Anthropic to offer access to its advanced AI system while limiting potential risks associated with sensitive subject areas.
The launch has generated mixed reactions among developers, researchers and AI enthusiasts. Supporters argue that Anthropic is taking responsible steps to address growing concerns about advanced AI systems and their potential misuse.
Others believe the restrictions are too broad and may reduce the model's usefulness for students, researchers and professionals seeking legitimate information.
Several users pointed out that refusing basic educational questions could frustrate people who simply want accurate scientific explanations.
Despite criticism over its safeguards, many users praised Claude Fable 5's technical performance. Early testers reported that the model performs exceptionally well during long sessions and demonstrates strong software development capabilities.
Some developers described the model as one of the most capable coding assistants released to date. At the same time, others expressed concern about how rapidly AI systems are improving.
One critic described the model as “by far the most ridiculous model that makes me genuinely afraid for the future of software engineering.”
The reaction to Claude Fable 5 highlights a growing challenge facing AI companies. As models become more capable, developers must balance usefulness with safety. Companies increasingly face pressure to prevent harmful uses of AI while still providing valuable tools for education, research and business.
Anthropic's decision to prioritise strict safeguards reflects that balancing act. Whether users ultimately accept those limitations may depend on how effectively the company can refine its safety systems without blocking ordinary questions.
For now, Claude Fable 5 has already achieved one thing: it has reignited the debate over how powerful AI models should be controlled as the technology continues to evolve.
