Silicon Valley's capture of our political institutions is all but complete (www.bloodinthemachine.com)

California Gov. Gavin Newsom’s recent actions on AI legislation highlight deep regulatory capture by Big Tech: he signed a few headline-grabbing but effectively toothless laws while vetoing tougher protections that tech lobbyists opposed. Wins included AB 325 (limits algorithmic dynamic pricing). Vetoes included SB 7 (No Robo Bosses Act), which would have barred employers from using purely automated systems to discipline or fire workers, and AB 1064 (LEAD Act), which would have required firms to ensure chatbots don’t harm minors. Meanwhile, Newsom signed SB 53, a “transparency” law forcing large AI firms (>$500M revenue) to self-report safety protocols and “catastrophic risks” (defined as events causing 50+ deaths or $1B damage) — violation penalties top out at $1M — and SB 243, which only requires firms publish suicide/self-harm query protocols and exposes them to trivial $1,000 civil damages. Why this matters: California often sets de facto national standards, so weak, self-reporting rules and tiny penalties institutionalize performative compliance rather than enforceable safety. Key technical implications include reliance on vendor self-assessment instead of independent audits, narrow and impractical “catastrophic” thresholds, inadequate whistleblower protections, and a missed chance to regulate high-risk uses (child-facing chatbots, automated employment decisions). The outcome risks emboldening AI firms to prioritize product rollout over robust safety engineering and oversight, making meaningful governance harder to achieve elsewhere.