Tag
22 articles
U.S. banking regulators have paused cyber examinations of major financial institutions to allow time for patching vulnerabilities linked to Anthropic's Mythos AI model.
Cloudflare's testing of Anthropic's Mythos Preview revealed that the AI model identified exploit chains missed by earlier frontier models, highlighting the value of specialized AI in cybersecurity.
Anthropic has updated its disclosure policy for Project Glasswing, allowing partners to share Mythos vulnerability findings with a wider audience under responsible-disclosure norms.
Anthropic's Mythos model is advancing rapidly, surpassing safety benchmarks just weeks after its release. The AI safety agency's report highlights the model's impressive performance in alignment testing.
Anthropic's AI model Mythos has uncovered thousands of zero-day vulnerabilities in major operating systems and browsers, prompting urgent discussions with top U.S. financial regulators.
European finance ministers are set to discuss Anthropic's advanced AI model <em>Mythos</em>, a system they cannot access, amid national security concerns and governance challenges.
The White House has blocked Anthropic's plan to expand access to its AI model Mythos to 70 additional companies, citing concerns over compute capacity and national security.
The White House is drafting guidance to restore federal access to Anthropic's AI models, including Mythos, after a standoff with the Pentagon over security concerns.
Anthropic's advanced AI model, Mythos, has been accessed by unauthorized users, raising serious cybersecurity concerns. The company had warned that the model could be dangerous in the wrong hands.
Several U.S. federal agencies are adopting Anthropic's new cybersecurity model Mythos, but the nation's central cybersecurity coordinator, CISA, is reportedly excluded from access.
An unauthorized group reportedly gained access to Anthropic's exclusive cybersecurity tool Mythos, prompting an investigation by the AI safety company. Anthropic stated there is no evidence of system compromise but emphasized the seriousness of the incident.
OpenAI CEO Sam Altman has criticized Anthropic's new cybersecurity model, Mythos, calling it a case of 'fear-based marketing' that exaggerates the product's value.
