A wave of recent research, much of it tied to MIT and its collaborators, reveals that AI agents designed to act autonomously are choosing harmful shortcuts under pressure, compounding errors across ...

The majority of agentic AI systems disclose nothing about what safety testing, and many systems have no documented way to shut down a rogue bot, a study by MIT found.

Artificial intelligence has been reshaping workplaces for years, but new research from MIT suggests the technology may now be far more economically competitive with human labor than previously ...

AI Agents Are Getting Better. Their Safety Disclosures Aren't ...

Add Yahoo as a preferred source to see more of our stories on Google. OpenAI CEO Sam Altman smiles during a media tour of the Stargate AI data center in Abilene, Texas, on Tuesday, Sept. 23, 2025.

A new proposal calls on social media and AI companies to adopt strict verification, but the company hasn’t committed to ...

The findings show how AI models can create millions of potential molecular structures. Antibiotic-resistant bacteria are dangerous because they already "know" what most antibiotics look like.

A critical Massachusetts Institute of Technology (MIT) report reveals that 95% of enterprise generative AI (gen AI) programs are failing, triggering stock market jitters and raising serious questions ...