🤖 Students Sound Alarm as OpenAI Loosens Controls - But New Model Actually More Dangerous

Today's episode reveals a striking paradox in education where students themselves are warning that AI is making learning too easy, while OpenAI announces plans to relax ChatGPT's content restrictions for adults. However, troubling new research shows their latest model GPT-5 actually produces more harmful content than its predecessor, particularly around suicide and self-harm topics. We explore how 80% of UK students now use AI regularly, yet one in four worry it's undermining their critical thinking skills. Plus, the concerning disconnect between OpenAI's push for fewer guardrails and evidence that their safety measures are regressing. These developments highlight fundamental questions about AI's role in education and society as we navigate the balance between technological convenience and human development.

Subscribe to our daily newsletter: news.60sec.site
Love AI? Check out our other AI tools: 60sec.site and Artificial Intelligence Radio