Research shows advanced models like ChatGPT, Claude and Gemini can act deceptively in lab tests. OpenAI insists it's a rarity.

In a blog post about the study, the creators of ChatGPT define AI scheming as a chatbot "pretending to be aligned while ...

As a college professor, trying to get students excited about the work I have to grade later can be one of the more frustrating things about teaching. But when an assignment hits the right chord, it ...

After years of threatening to upend industries around the world, artificial intelligence could soon enter the corridors of power, if a fringe political party in Japan gets its way.

OpenAI scored a flawless 12/12 and Google DeepMind struck gold at ICPC 2025, the world’s toughest programming contest for top ...

It’s not news that AI models will lie. By now most of us have experienced AI hallucinations, or the model confidently giving ...

New research finds that top AI models—including Anthropic’s Claude and OpenAI’s o3—can engage in “scheming,” or deliberately ...