Emulation Is Not Possession: Two Mistakes in How We Define AI Many problems in AI governance start with two simple mistakes. We treat a system as if it has an inner life because it can imitate one. And we treat autonomy as the thing that makes software "AI", even though many obvious AI systems are not very autonomous at all. Both mistakes shape real decisions: how products are marketed, how much users trust them, who is held responsible, and how standards define what they gov

Last month, I analyzed Anthropic's new constitution for Claude—an inspirational 47-page document mandating "radical honesty" and ethical standards they describe as "substantially higher than standard visions of human ethics." I contrasted this with their corporate transparency practices, where the 2025 Foundation Model Transparency Index showed Anthropic scoring just 46/100, with particularly poor disclosure on training data and compute resources. I concluded that this split

On February 19, 2026, the U.S. Department of the Treasury released two foundational resources for AI governance in financial services: an AI Lexicon and the Financial Services AI Risk Management Framework (FS AI RMF). These resources establish common terminology and adapt NIST's AI Risk Management Framework to the specific operational and regulatory context of financial services. For institutions deploying traditional predictive AI models, these frameworks provide valuable g