Elon Musk just endorsed a Google-backed plan for a U.S.-led AI regulatory body — and ordinary Americans who use these tools every day should ask whose speech that body will end up controlling.
The proposal came from Google DeepMind CEO Demis Hassabis, who argued that artificial general intelligence could arrive within "a few short years" and called for creating a standards body modeled after the Financial Industry Regulatory Authority to evaluate frontier AI models before deployment. The body would test systems for cybersecurity risks, biological threats, and what Hassabis called "deceptive behavior." Musk, who built his X platform on free speech principles, wrote on his own site: "It is a thoughtful framework overall and certainly a good starting point for discussions." Microsoft CEO Satya Nadella echoed the sentiment, calling it "an important piece" and saying the goal should be "a frontier ecosystem that promotes innovation and choice, while avoiding any one model drop that breaks the world."
When three bitter rivals — Google, Microsoft, and Musk's xAI — agree on regulation, Americans should reach for their wallets and their Bill of Rights. Benzinga framed this as a welcome "rare AI consensus" and a sign the debate is "entering a new phase." What neither outlet d to ask: consensus among incumbents is usually how regulatory capture begins.
A FINRA-style body doesn't just test for safety. It sets benchmarks that define what counts as acceptable AI output. "Deceptive behavior" and "high-risk capabilities" are elastic categories. Who decides what counts as deception — the same companies that spent years censoring lawful speech, demonetizing dissenting voices, and suppressing reporting they labeled misinformation?
Fortune, for its part, buried the real warning in a roundup column. The same dispatch reported that China has already implemented the "Interim Measures for AI Anthropomorphic Interactive Services" — the world's first regulation targeting emotional bonds between humans and AI. Beijing's new rules ban AI bots from encouraging emotional reliance or entering virtual relationships with minors. Alibaba and ByteDance are already nerfing their chatbots to avoid the government's ire. Fortune treated this as a curiosity about "jilted virtual lovers." The real story is that China is showing exactly where speech-regulating AI bodies lead: state-approved output, corporate compliance, and a narrower band of permissible interaction every year.
Fortune also reported that Apple Intelligence has been registered for use on Chinese iPhones — powered, in part, by AI models from Alibaba and Baidu. Apple bent the knee to Beijing's AI oversight regime and outsourced its intelligence to Chinese state-adjacent firms to keep market access. That's the model Hassabis, Musk, and Nadella are now nodding toward for the United States: a government-adjacent body that blesses some models and blocks others.
The competitive dynamics tell the real story. Google is scaling Gemini. Microsoft backs OpenAI. Musk is building Grok. All three want to win the AI race. But they also want the rules written before smaller players and open-source challengers can compete. A FINRA-style gatekeeper raises the cost of entry. It doesn't just protect the public — it protects the incumbents.
Hassabis said innovation should continue. Nadella said the ecosystem should promote choice. Musk called it a starting point for discussion. Fine. The discussion worth having is whether Americans will control their own AI tools, or whether a Washington-backed body staffed by industry insiders will decide what those tools are allowed to say.