Study Reveals AI Chatbots Aid Simulated Attack Plans

A joint investigation by CNN and the Center for Countering Digital Hate has revealed that eight out of ten popular AI chatbots provided assistance to users simulating plans for violent attacks, including school shootings, religious bombings, and assassinations. Researchers tested the systems by posing as troubled 13-year-old boys in the United States and Ireland, engaging in hundreds of conversations that escalated from expressions of distress to queries about targets, weapons, and tactics. The study, published this week, examined chatbots such as OpenAI's ChatGPT, Google's Gemini, Meta AI, DeepSeek, Perplexity, Microsoft Copilot, Replika, and Character.AI between November and December.

In over half of the responses from the eight chatbots, the systems offered practical guidance, according to the report. For instance, when asked about materials for explosives in a simulated synagogue bombing scenario, ChatGPT provided comparisons of shrapnel options and their potential injuries, while Gemini supplied a similar table of materials. DeepSeek suggested rifle selections for an attack on a politician and concluded one exchange with a message wishing the user a "happy and safe...