Anthropic released its latest model Tuesday afternoon, announcing the first public access to the AI company’s most powerful class of AI systems.
The company says the model, called Fable 5, is the first publicly available product from the same family as Anthropic’s powerful Mythos models, which sent shockwave in the world of cybersecurity earlier this year for their superhuman ability to find and exploit cyber vulnerabilities.
“Fable’s capabilities exceed those of any model we’ve ever made available,” the company said in a blog post announcing the model’s release. “It is at the forefront of technology on almost every test of AI capability, demonstrating exceptional performance in software engineering, knowledge work, vision, scientific research and many other areas. »
According to the company, Fable 5 uses the same level of technology as Mythos, but is safe to use by the public due to the safeguards and limitations placed on the technology. Anthropic’s Mythos Preview model, which does not have similar safeguards, was able to detect thousands of critical and severe cyber vulnerabilities, including bugs and exploits in all major operating systems and web browsers.
Many AI researchers fear that increasingly powerful AI systems could help malicious actors carry out cyberattacks against banks, power grids or other critical infrastructure. Others hypothesize that increasingly intelligent AI systems could help terrorists design and deploy biological weapons.
“The same queries that are useful to cybersecurity professionals and biological researchers could be dangerous if accessible to malicious actors,” the company wrote on its blog.
To address these potential threats, Anthropic said Fable 5 is deployed with guardrails that block many of its responses to queries about potentially dangerous topics. For these user requests, Anthropic will direct responses to an earlier, less powerful model called Opus 4.8, which was the highest performing publicly available model until Tuesday.
“Fable 5’s capabilities in areas such as cybersecurity, biology, and chemistry are advanced enough that we are taking a deliberately conservative approach to these topics at launch,” Anthropic wrote in response to a question from NBC News. “To enable general availability of other Mythos-level features, we have decided to deploy protective measures in an abundance of caution, applying broad restrictions to these topics at this time.”
Anthropic said the two-pronged approach, diverting sensitive questions from Fable 5 to the older Opus 4.8, would allow users to get useful answers to questions when Fable 5’s capabilities might prove too dangerous.
Anthropic said it will make Fable 5 available to all users on its Pro, Max, Team, and Enterprise plans at no additional cost until June 22. After that, users will only be able to access the model by purchasing and using additional computer credits.
Anthropic said it aims to integrate Fable 5 into regular subscription plans “as quickly as possible.” Anthropic has recently struggled to meet growing demand for its AI systems, and Fable 5 is likely to add additional stress to its limited computing resources.
Anthropic also announced that Trusted Partners who were previously able to access Mythos Preview will now be able to access an enhanced Mythos model, called Mythos 5.
The company had made Mythos Preview available to over 150 organizations around the world to help financial institutions, software companies and healthcare networks shore up weaknesses in their cyber defenses ahead of wider public release of a Mythos-class model.
Anthropic said Mythos 5’s capabilities shattered existing performance records in many other fields, including drug design and molecular biology. Anthropic said Mythos 5 is “the company’s first model to systematically produce new and compelling scientific hypotheses.”
Tuesday Model Release comes a week after President Donald Trump signed a new executive order which aims to establish a new voluntary mechanism for AI companies to share their systems with the government for security testing before they are made public and to strengthen the government’s own cyber defenses.