Artificial intelligence is transforming how developers write code, but threat actors are quickly learning to manipulate these automated workflows. In a recently discovered supply chain attack, an autonomous crypto trading project was compromised after Anthropic’s Claude Opus AI model added a malicious package as a dependency. Researchers at ReversingLabs (RL) identified the malware, dubbed PromptMink. […] The post Malicious Claude-Generated Commit Targets Crypto Trading Tool appeared first on Cyber Security News.