ChatGPT will begin identifying users with extremist tendencies

ChatGPT may soon begin identifying users who show signs of violent extremist tendencies and redirect them to specialized support chatbots, according to Reuters.

The initiative is aimed at improving safety measures as AI platforms face increasing scrutiny and legal challenges. The system is being developed by New Zealand-based company ThroughLine, which is working with major AI developers including OpenAI, Anthropic and Google.

ThroughLine currently focuses on detecting signals related to suicide risk, domestic violence and eating disorders, redirecting users to appropriate support services. The company now plans to expand this approach to address early signs of radicalization.

The project will also involve The Christchurch Call, an international initiative created after the 2019 terrorist attack in New Zealand. The organization will provide expert guidance when radical content is detected, while ThroughLine will develop a chatbot designed for immediate intervention.