These AI Experts Want to Do High-Stakes Research Out in the Open

# AI Lab Takes Unusual Approach to Risky Research Most AI companies keep their most dangerous experiments hidden from the public, but Trillium Labs is doing the opposite—they want to openly share their work on how AI systems improve themselves and behave. The lab believes that transparency about high-stakes AI research will actually make these systems safer, rather than putting people at risk. It's a bold bet that public scrutiny and collaboration beat secrecy when it comes to preventing AI problems down the line.