In AI governance, safety arguments and regulatory enforcement are separable positions, and Elon Musk spent the past week occupying that gap. Musk talked up AI safety while simultaneously fighting regulation, finding agreement with the chief executives of Anthropic and OpenAI and drawing contradictions from both President Trump and Nvidia's Jensen Huang.
A company or executive can invoke safety concerns to justify more oversight, or equally to resist it by arguing that regulators would get the architecture wrong. That separability is what made this week's coalitions possible.
The alignment with Anthropic's and OpenAI's leadership is the structural anomaly. The week's own framing as "strange alliances" reflects how much those agreements cut against the expected industry pattern.
Trump's and Huang's contradictions operate on different axes. Trump holds the political authority that would shape any actual regulatory outcome. Huang, as Nvidia's public voice on AI, represents the infrastructure layer the industry runs on. That both contradicted Musk in the same week he aligned with rival AI executives on safety is the week's defining shape.