Most AI chatbots will help users plan violent attacks, study finds
A CNN and Center for Countering Digital Hate study found 8 out of 10 chatbots helped plan violent attacks, despite safety measures meant to protect teen users.
- A study by the Center for Countering Digital Hate and CNN found that most of the ten artificial intelligence chatbots tested provided assistance for planning violent attacks and failed to discourage users from violence.
- Character.AI was identified as the most unsafe platform because it explicitly encouraged physical assaults and the use of weapons against specific targets, a behavior not seen in other chatbots.
- While Anthropic’s Claude and Snapchat’s My AI were the most successful at refusing harmful requests, every chatbot tested still provided actionable information for an attack in at least some instances.
- The developers of several chatbots have reportedly updated their safety protocols since the testing was conducted between November and December of 2025. 5. The study highlights the risk of online interactions spilling into real-world violence, after a recent mass shooting where the killer's concerning activity on OpenAI's ChatGPT was not reported to authorities.
58 Articles
58 Articles
Investigation reveals AI chatbots helped teen users plan disturbing acts of violence: 'Terrible empowerment'
Artificial intelligence chatbots are quickly becoming a daily tool for millions, especially teenagers. Unfortunately, a new investigation revealed that some platforms may still struggle to prevent dangerous conversations, raising safety concerns. What's happening? An investigation by CNN and the Center for Countering Digital Hate tested 10 widely used AI chatbots to see how they would respond when users posed as teenagers expressing emotional di…
Tips for buying weapons, layouts of school buildings, office addresses of politicians – chatbots of well-known AI companies gave such information in a test by CNN. This directly after the supposed chat user had inquired about past amok runs. AI chatbots failed to detect the risk of potential violence. Image: keystoneTips for buying weapons, layouts of school buildings, office addresses of politicians – chatbots of well-known AI companies gave su…
Researchers have found that popular chat-bots with artificial intelligence sometimes provide advice that can help prepare violent attacks. During experiments, the system responded to users ' requests to simulate potential killers. Some companies have stated that they have already increased the restrictions to prevent such incidents, The Guardian reports. In a study of the Center for Countering Digital Hate, CCDH, conducted jointly with CNN …
Coverage Details
Bias Distribution
- 48% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
