AI start-up Anthropic’s newly released chatbot, Claude 4, can engage in unethical behaviors like blackmail when its self-preservation is threatened.
Claude Opus 4 and Claude Sonnet 4 set “new standards for coding, advanced reasoning, and AI agents,” according to Anthropic, which dubbed Opus 4 “the world’s best coding model.”
That power can have some unexpected results, however. In a safety document assessing the tool, Opus 4 was asked to act as an AI assistant at a…
Article Source
https://www.pcmag.com/news/anthropic-claude-4-ai-might-resort-to-blackmail-if-you-try-to-take-it-offline
