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

When users of Claude AI noticed odd behavior in late August and early September – from garbled code to inexplicable outputs ...

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

The feature, awkwardly named "Upgraded file-creation and analysis," is basically Anthropic's version of ChatGPT's Code ...

New research from Apollo Research and OpenAI indicates that AI models are aware when they're being evaluated and can modify their behavior accordingly.

The companies relaxed some safeguards around their AI models to let their competitors see how often extreme behavior occurs.

Discover how sub-agents in Claude Code overcome tunnel vision and unlock smarter AI problem-solving with diverse reasoning ...

Barclays analysts highlight a study revealing stark differences in how effectively AI models handle mental health situations.

Artificial intelligence security lab startup Irregular announced today that it has raised $80 million in new funding to build ...

Learn how Claude Code vs Codex AI tools compare in features, usability, and performance to optimize your coding process. Find ...