Anthropic says it has blocked attempts by malicious actors to misuse its Claude AI models for cyberattacks, surveillance and research that could have led to biological weapons. Sky's Barnaby Papadopulos explains the company's findings, including attempts to use AI for influence operations, weapons-...