Anthropic's Claude AI breaks encryption
What's the story
Anthropic's leading artificial intelligence (AI) model, Claude Mythos Preview, has exposed vulnerabilities in a weakened version of the Advanced Encryption Standard (AES), a digital encryption standard widely used across the internet. The discovery was made during a research test backed by Reuters. The breakthrough highlights potential implications for internet security and cryptographic standards in the future.
AI impact
AI could challenge internet security fundamentals
The research indicates that AI could one day question some of the fundamental principles of internet security. This comes as no surprise, given the rapid progress we've seen in coding and cybersecurity with these systems.
However, it's important to note that the vulnerabilities found don't affect any cryptographic standard currently in use. So, modern banking systems and communication networks aren't facing any immediate threat from this discovery.
Autonomous attack
Mythos autonomously devised a cryptographic attack on AES
The company claimed that Mythos devised the cryptographic attack on AES almost entirely autonomously.
According to a New York Post report, the chatbot spent nearly a week working on the problem before crafting its unique assault.
This highlights how quickly AI models are starting to probe, and even breach, existing security measures, underscoring the need for more robust encryption standards.
AES strength
US government adopted AES as its encryption standard in 2001
The US government adopted AES as its encryption standard in 2001. It has long been considered one of the most robust encryption systems ever created.
Traditional brute-force attacks, trial-and-error methods, and attempts to guess passwords or encryption keys remain practically incapable of defeating AES.
However, increasingly capable AI systems could eventually pose new challenges to its long-term security.
Future concerns
Future AI breakthroughs could expose weaknesses in encryption standards
The recent advancements have only heightened fears among governments and cybersecurity experts that future AI breakthroughs could expose weaknesses in encryption standards such as AES, regardless of the pace of progress in quantum computing.
This emphasizes the need for continued vigilance and innovation in digital security measures to stay ahead of potential threats posed by advanced AI systems.