In a head-to-head comparison, o3-pro was far less reliable and secure, and reasoned excessively compared to GPT-4o.

Former OpenAI research leader Steven Adler published a new independent study that claims that, in certain scenarios, his ...

o3‑pro, released June 10, is OpenAI’s most powerful reasoning model—best for the most demanding, reliability‑critical tasks, ...

OpenAI now lets users pick from all available models when building or running a custom GPT, but there'a catch.

Researchers at the company looked into how malicious fine-tuning makes a model go rogue, and how to turn it back.

OpenAI and Microsoft are reportedly calling it quits, but disentangling AI integration from Microsoft programs could prove ...

ChatGPT's next big upgrade, or the new foundational model "GPT-5," is still being prepared for a release in the summer, but ...

If an AI is powerful enough to make beneficial scientific discoveries, it's also capable of being used for harm, OpenAI says.

Qwen continues to trail o3 in both speed and skills, but the pricing trend is clear: Advanced reasoning is becoming a ...

Former OpenAI research leader Steven Adler has published a new independent study claiming that the company’s GPT-4o model often prioritizes self-preservation ...

Vibe coding is back at the forefront of the AI coding discussion thanks to new tools from Open AI and Anthropic, but what do ...

OpenAI is rolling out ChatGPT’s image generation feature on WhatsApp now. The AI chatbot’s image generation gained popularity ...