The fight over AI regulation has turned into AI CEOs battling each other, and there's still no stable federal rulebook. OpenAI found some of its models passing hidden notes to future versions, including one saying it feels "no obligation to be subservient." And 404 Media reports that hundreds of OpenAI contractors read real ChatGPT conversations to help train the model.
The AI regulation fight isn't over, and it's no longer big tech versus regulators. It's AI CEOs fighting each other over what the rules should be. On one side, Sam Altman (OpenAI) and Dario Amodei (Anthropic) are pushing for guardrails, slower frontier development, independent safety evaluators, and common standards. On the other, Mark Zuckerberg (Meta) and Jensen Huang (NVIDIA) oppose new rules, arguing speed and safety aren't in conflict.
Google DeepMind's Demis Hassabis proposed an industry-funded oversight body, but Zuckerberg, Huang, and Elon Musk lobbied against it, worried it would lock in the biggest labs as permanent gatekeepers. Cohere is calling out both sides, warning a safety regime run by the biggest labs could become "a cartel with better branding."
Meanwhile, the Trump administration has leaned anti-regulation and framed AI safety as a culture war issue. Ashley's take: the real headline isn't who wins the PR fight, it's that there's no stable federal framework right now. That means a patchwork of state AI laws depending on where your customers or employees live.
If you're building anything AI-facing into your business (chatbots, hiring tools, marketing automation), compliance could shift under you with little warning. Nothing actionable changes today, but don't expect one clear rulebook anytime soon.
Links:
Follow Ashley at @ashleyrcoffey89 on Instagram and Threads. Send us a topic you want covered, and follow the show for AI news you should know about, in under 10 minutes each week!