A major breakthrough in cybersecurity: Anthropic's AI model Claude has identified critical vulnerabilities in two encryption systems. This concerns the mathematical algorithms that underpin the protection of messages, logins, and banking data from unauthorized access.

It is important to emphasize: current encryption standards used everywhere are not yet in danger. However, the significance of this discovery cannot be overstated—the machine detected what leading global experts, who had worked on these algorithms for years, had overlooked.

HAWK: A Blow to Post-Quantum Hope

The HAWK system is a digital signature designed with resistance to quantum computer attacks in mind. Over two years of research, experts found no flaws in it. Claude needed only 60 hours to notice a hidden mathematical pattern.

According to calculations, cracking the smallest HAWK key is now possible in 2^38 steps instead of 2^64. This means a reduction in computational costs by approximately 67 million times. A solution exists, but it is extremely unpleasant: the key size must be doubled, negating HAWK's main advantage—its efficiency.

AES: "Möbius Bridge" Accelerates Attack

The AES standard, which has been the foundation of data protection on the internet since 2001, also underwent analysis. Full AES remained intact, but Claude examined its simplified version with 7 rounds instead of 10. The AI found a non-standard attack method, dubbed the "Möbius Bridge," which accelerated key recovery by 200–800 times.

Notably, Claude worked alone for three days, and then two specialists needed nearly a month to verify its results. This raises a fundamental question: should the security checks of our data be entirely entrusted to artificial intelligence?

Insights from Cryptalist Expert

This event is not just another news item but a wake-up call for the entire crypto industry. Although Bitcoin and Ethereum use different principles and are not yet affected, time is not on our side. NIST will need to evaluate the public attack on an algorithm that has not yet been adopted as a standard. Anthropic has already hinted that not all results have been published and is preparing a special seminar.

My professional opinion: We stand on the threshold of a new era where AI becomes not just a tool but the primary security auditor. The quantum deadline of 2031 for U.S. government agencies and Google's plan to migrate by 2029 are not just dates but real timelines within which the industry must rethink its approaches. The vulnerabilities found by Claude are just the tip of the iceberg. Investors and developers should seriously consider the post-quantum resilience of their assets today, rather than waiting for "Q-day" to arrive.