Anthropic Launches Mythos-Class Models After Developing Safeguards

Claude Mythos

Anthropic has launched two Mythos-class models after developing safeguards to prevent them from being misused, the company said in a Tuesday (June 9) press release.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    Claude Fable 5 is safe for general use, while Claude Mythos 5 will initially be released for only a small group of cyber defenders and infrastructure providers through Project Glasswing, according to the release.

    The launch of these models follows the limited release in April of the first Mythos-class model, Claude Mythos Preview. Anthropic introduced Project Glasswing at that time, saying it would offer select partners early access to Claude Mythos Preview so they could use the model’s cybersecurity capabilities to strengthen their systems.

    By May 22, Anthropic reported that Claude Mythos Preview had identified more than 10,000 cybersecurity vulnerabilities in “the most systemically important software in the world” so that they could be patched.

    Anthropic said in the Tuesday press release that since April, it has been developing safeguards for Mythos-class models.

    “Over the past few months we have been improving these safeguards, and they are now robust enough for a general release,” Anthropic said in the Tuesday press release. The company said that it has prioritized safety, so users may encounter false positives from the safeguards.

    Claude Fable 5 has capabilities that exceed those of any previous model Anthropic has made generally available, is state-of-the-art on most benchmarks of AI capability, and performs especially well in software engineering, knowledge work, vision and scientific research, according to the release.

    Because Fable 5’s capabilities could be misused, it has safeguards that will route queries on some topics to the less-capable Claude Opus 4.8. Those topics include cybersecurity, biology and chemistry, and distillation, per the release.

    Claude Mythos 5 is the same underlying model as Fable 5 but with cyber safeguards lifted, according to the release. Anthropic said Mythos 5 has “the strongest cybersecurity capabilities of any model in the world” and will initially be deployed through Project Glasswing in collaboration with the U.S. government. The company plans to expand access through a trusted access program.

    Mythos 5 was made available Tuesday to users who have access to Claude Mythos Preview. It is comparable to, or somewhat stronger than, that model but costs substantially less.

    “The capabilities of models like Fable 5 and Mythos 5 have the potential to do profound good for the world,” Anthropic said in the release. “We’ve seen the beginnings of this in Project Glasswing, where the models have helped cyber defenders secure critically important software. We’ve also seen it in life sciences research, where the models are positing novel hypotheses and speeding up the development of new therapeutics.”