New Zealand Develops Tool to Counter Extremism in AI Chatbots
Technology firms and experts in New Zealand are developing a tool aimed at tackling violent extremism on artificial intelligence platforms. The project seeks to redirect at-risk users to support combining expert intervention with chatbot assistance.
The initiative is being developed by startup ThroughLine, which has worked in recent years with ChatGPT owner OpenAI as well as competitors Anthropic and Google. The company is also in talks with the Christchurch Call, an initiative launched after the 2019 terror attack to curb online hate.
According to the developers, the system would operate as a hybrid model. A chatbot would respond to warning signs in conversations while directing users to support services in their area. The developers said the solution remains in testing and no deployment date has been set.
Experts say the growing popularity of chatbots has widened the range of issues people share online, including extremist views. They are also debating how follow-up measures, such as involving authorities, could be implemented without risking further escalation.