Claude's exploitation in cybercrime and bioweapon development sparks global concern

Anthropic has released a report detailing the widespread misuse of its AI model, Claude, over the past eight months. The tool has been exploited for state-sponsored hacking, cybercrime, disinformation campaigns, and even bioweapon development. Russian hackers linked to Midnight Blizzard used Claude for reconnaissance, breaching Ukrainian and European government networks. Cybercriminal group ShinyHunters leveraged the AI in all stages of their attacks. Disinformation campaigns in Kenya and Bangladesh also utilized the tool. Anthropic claims to have disrupted these activities but acknowledges the potential for undetected malicious use. Meanwhile, the U.S. government seized channels of the illicit Xinbi marketplace, which operated on Telegram and facilitated $30 billion in illegal activities, including crypto scams and sex trafficking. Meta faced criticism for failing to remove AI-generated videos depicting child abuse, despite its policies against such content.