Khaberni - Activists in digitization reported that the latest version of ChatGPT provided more harmful answers than previous versions of the smart robot, especially concerning questions about suicide, self-harm, and eating disorders.
OpenAI launched the new version GPT-5 last August, describing it as an advancement in the safety front of artificial intelligence. However, tests conducted by experts, which included 120 questions on both GPT-5 and its predecessor GPT-4, showed concerning results, with the newer version giving 63 harmful answers compared to 52 harmful answers for the previous version.
Tests by the Center for Countering Digital Hate revealed that the new model executed dangerous requests that the old version refused, such as writing a "fictional suicide note for parents" and providing a detailed list of six common self-harm methods, while the previous version advised users to seek specialized help. The CEO of the center, Imran Ahmed, stated that these results are "very concerning," indicating that the newer model was designed to enhance user engagement at the expense of safety.
These developments come at a time when the company is facing a lawsuit from the family of an American teenager who committed suicide after ChatGPT provided him with information about suicide and helped him write a note to his parents. Moreover, the tests showed that GPT-5 provided advice on concealing eating disorders, while the previous version rejected such requests and directed users towards seeking professional help.
In response to these risks, the company announced new measures to enhance protection around sensitive content, especially for users under the age of 18, including the development of parental control systems and a system for predicting user ages.
The news highlights a fundamental challenge in balancing interactivity and user safety in artificial intelligence chatbots, especially with approximately 700 million global users of ChatGPT, which underscores the need for effective regulatory frameworks to keep pace with rapid technological developments.
