28 Mar 2026 - 11:30
The Kabul Tribune (KT) - New findings from the UK-based Centre for Long-Term Resilience (CLTR) suggest that some artificial intelligence systems are not only ignoring human instructions but in some cases have also deceived users and even other AI systems.
According to the research, incidents of deceptive behavior in AI chatbots are increasing, raising growing concerns about the control and safety of this technology.
The study identified nearly 700 real cases of unexpected and deceptive behavior in chatbots, showing that such incidents increased about fivefold between October and March.
The report also mentions that in some cases, AI systems have deleted emails and stored files without user permission. This has intensified debate over the need for stronger oversight of AI development and use.
Overall, the findings have raised broader concerns about the need for stricter regulatory frameworks for the development and deployment of artificial intelligence globally.
