Anthropic's artificial intelligence Claude has achieved what leading cryptographers worldwide failed to do for years. During research, the model discovered critical vulnerabilities in two encryption systems — HAWK and a simplified version of AES. These are mathematical algorithms that underpin the protection of messages, logins, banking data, and many other confidential pieces of information.
It is important to emphasize: current encryption standards used by millions of people are not yet under threat. However, the very fact that a machine was able to find what experts missed over years of work changes the rules of the game in the field of cybersecurity.
HAWK: Vulnerability Discovered in 60 Hours
The HAWK digital signature system was originally designed with resistance to quantum computing in mind. For two years, specialists found no flaws in it. But Claude needed only 60 hours to identify a hidden mathematical pattern at the core of the algorithm.
According to Anthropic, 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. There is a solution to the problem, but it is extremely unpleasant: the key size must be doubled, which effectively nullifies the system's main advantage.
AES: The Möbius Bridge Attack
AES (Advanced Encryption Standard) is the primary standard protecting most internet traffic since 2001. Claude could not crack full AES but analyzed a simplified version with 7 rounds instead of the standard 10.
The model discovered a non-standard attack method, named Möbius Bridge. The essence of the trick is that one stage of the selection process turned out to be redundant. As a result, the most effective attack was accelerated by 200 to 800 times.
Notably, Claude worked alone for three days, after which two specialists needed almost a month to double-check the results.
What This Means for the Crypto Industry
The main delay now occurs at the verification stage. Anthropic notes: humans can no longer keep up with the model's pace of work. This is not the first time vulnerabilities have been found only at late stages. In 2022, researchers managed to crack another post-quantum algorithm, SIKE, in about an hour on an ordinary laptop.
Bitcoin (BTC) and Ethereum (ETH) operate on a different principle, and these attacks do not affect them. However, time does not stand still. Google plans to transition to new standards by 2029, and U.S. government agencies have a quantum deadline of 2031.
Now NIST will have to evaluate a public attack on an algorithm that was still being considered for implementation. Anthropic has indicated that not all results have been published yet, and the company will soon hold a special seminar with scientists.
My professional opinion: This case is a clear signal for the entire crypto industry. AI not only accelerates analysis but also finds non-obvious connections that humans might miss for years. The question is not whether AI can crack current standards, but whether we can adapt to a new reality where machines check security faster than humans can create it.