Base76 research lab

Tag: AI governance

  • TrustPlane – our latest safety product for AI

    Every unverified AI response in production is an undocumented decision There is no entry in your log showing what the model was asked. No entry showing how confident it was. No entry showing whether anyone reviewed the response before it reached your user. In most industries, that is acceptable. In healthcare, legal, finance, and public…

  • Anthropic vs DoD — An Epistemic Governance Analysis

    By: Bjorn WIkstrom Executive Summary The emerging conflict between Anthropic and the U.S. Department of Defense (DoD) is not primarily about defense contracts or AI ethics rhetoric. It is a structural dispute about epistemic governance authority: Who has the legitimate right to determine how uncertainty, capability, and risk are interpreted inside frontier AI systems? This…