Anthropic's artificial intelligence model, Claude, has achieved what leading cryptographers worldwide couldn't accomplish for two years. In just 60 hours of work, the model identified hidden vulnerabilities in two encryption systems — HAWK and a simplified version of AES. These involve the mathematical methods underlying the protection of messages, logins, and banking data.

What's wrong with HAWK?

HAWK is a digital signature designed with future threats from quantum computers in mind. Experts spent two years searching for flaws — and found none. Claude discovered a hidden pattern in HAWK's mathematical foundation in just 60 hours. 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 cost by approximately 67 million times. A solution exists, but it's unpleasant: the key size must be doubled, which virtually nullifies the algorithm's main advantage.

Breakthrough in AES analysis

AES is the primary encryption standard protecting most internet traffic since 2001. Claude couldn't crack full AES but analyzed a simplified version with 7 rounds instead of the standard 10. The model found an unconventional approach, named Möbius Bridge, which accelerated the most effective attack by 200–800 times. Claude worked alone for three days. Then two specialists spent nearly a month double-checking the results.

Why is this important?

The main delay now occurs at the verification stage. Humans can no longer keep up with the model's pace. Similar cases have occurred before — in 2022, researchers cracked another post-quantum algorithm, SIKE, in about an hour on a regular laptop. However, the situation is different now: AI operates autonomously and faster.

Bitcoin (BTC) and Ethereum (ETH) operate on a different principle, so these attacks do not affect them. But time is not on our side. Google plans to transition to new standards by 2029, and U.S. government agencies have a quantum deadline of 2031.

NIST now must evaluate a public attack on an algorithm still under consideration for implementation. Anthropic has indicated that not all attack results have been published yet. The company plans to hold a special seminar with scientists in the near future.

Expert opinion: This case is not just a technical novelty. It marks the beginning of a new era where AI becomes the primary tool for security auditing. The question is not whether AI can find a vulnerability, but whether we can verify its findings before they are exploited by malicious actors. The crypto industry should prepare for accelerated standard update cycles.