Anthropic applies AI to detect vulnerabilities in blockchain smart contracts
Anthropic has started using its large language models (LLM) to identify vulnerabilities in blockchain smart contracts. Their model has already discovered an exploit capable of causing approximately $4.6 million in losses.
Source Anthropic
Analysts emphasize that artificial intelligence has significant potential for use in hacker attacks within the crypto industry. In simulations of real smart contract hacks from 2020 to 2025, systems such as Claude Opus 4.5, Sonnet 4.5, and GPT-5 demonstrated high effectiveness.
This development opens new opportunities for blockchain security but also increases cybersecurity risks, highlighting the need for enhanced protection measures and proactive counteractions.
