Anthropic made a new version of its Claude artificial intelligence technology available to the general public on Tuesday, while limiting its use in sensitive areas. Dubbed Fable 5, the model is the first wide release from the company's Mythos class—its most advanced AI lineup, unveiled in April but restricted to partner institutions for months due to cybersecurity worries.
The move comes as Anthropic navigates tensions between innovation and safety. The company had kept Fable 5's more powerful sibling, Fable 5's full version, under wraps after internal assessments flagged potential risks. Critics argue that the restricted release may still pose dangers, as even a 'safe' version could be exploited.
According to Reuters, the public version of Mythos lacks certain cybersecurity capabilities that were present in the original. The Guardian reported that access to the most powerful Mythos model, Fable 5, was restricted over cybersecurity concerns for months. Anthropic has not disclosed specific metrics on the model's performance or risk profile.
Microsoft AI head called out Anthropic for acting like Claude is conscious, according to The Verge, a claim that Anthropic has not publicly addressed. The broader industry is watching closely, as rivals like OpenAI and Google face similar scrutiny over the safety of their own advanced models.
The public release could accelerate AI adoption but also reignite debates around regulation. Some experts say voluntary safeguards may not be enough to prevent misuse, given the rapid pace of deployment.