In recent months, Anthropic has introduced Claude Fable, a new AI model designed with stringent safeguards to prevent misuse in biological research. The company, based in the United States, has stated that these measures are in place to protect against potential bioweapons threats. As a result, Fable is unable to answer even basic biology questions, such as those related to prions or mRNA vaccines, which are typically expected knowledge for high school students. This decision reflects Anthropic's commitment to safety, prioritizing the prevention of misuse over the model's ability to provide comprehensive biological information. The restrictions imposed on Claude Fable are part of a broader strategy by Anthropic to mitigate risks associated with advanced AI technologies. The company has identified four key areas where it will limit responses: chemistry, biology, cybersecurity, and distillation. These areas are considered high-risk due to their potential applications in harmful activities. For instance, while Fable can provide information on the use of chlorine gas as a chemical weapon, it refrains from offering synthesis instructions for dangerous substances like anthrax. This cautious approach aims to balance the benefits of AI capabilities with the need for safety and ethical considerations. Anthropic's spokesperson, Paruul Maheshwary, emphasized that the decision to implement these safeguards was made to allow customers to access the model's capabilities sooner while minimizing risks. The company is actively working to improve its detection systems and reduce false positives in the future. Despite the limitations, Anthropic believes that the launch of Claude Fable 5, its first Mythos-class model, represents a significant advancement in AI technology, with the potential to contribute positively to scientific research and drug discovery. Looking ahead, Anthropic has indicated that it intends to make Mythos-class models available without these stringent safeguards to the broader biology and life sciences community. This move is aimed at accelerating biomedical research and drug discovery, allowing researchers to leverage the capabilities of AI in a safe and responsible manner. However, the company has not clarified whether this restricted release will become a standard practice for future models, leaving the future of AI in sensitive fields uncertain.