A Brief History of AI Executives Calling for Regulation

The Industry's Long-Standing Calls for Regulation

For years, leading AI executives have publicly welcomed the idea of legal guardrails around their own products. In 2023, Sam Altman told a Senate Judiciary subcommittee that he supported licensing and safety requirements for advanced AI models, and OpenAI has since published frameworks calling for government oversight. Anthropic, Google DeepMind, and Microsoft have made similar statements, and hundreds of industry figures have signed open letters warning about existential risk.

That history makes the current backlash striking. The same companies that spent years asking lawmakers to act are now fighting the rules that followed. OpenAI and Anthropic have lobbied against state-level AI safety bills, while industry groups have challenged transparency requirements in court.

The pattern is not new. As one observer put it, the industry's enthusiasm for regulation tends to last right up until the regulation is written. What executives described as a shared responsibility now looks, from the outside, like a strategy: shape the rules before someone else does, then resist the ones that actually bind.

Why the Panic Now?

For years, established industry players have publicly welcomed regulation, insisting that clear rules would benefit everyone. So why the sudden alarm? The answer lies in what the current proposals actually contain. Earlier calls for regulation were often vague endorsements of "sensible" frameworks — the kind that incumbents could help shape and comfortably absorb. The rules now on the table are different: they target the very practices that have made these companies so dominant.

What looks like principle quickly reveals itself as positioning. When an industry asks to be regulated, it is often betting that it will write the rules or that the rules will fall hardest on smaller challengers. That bet no longer looks safe. The panic, then, is not a change of heart about regulation itself. It is the discovery that regulation might finally apply to them.

The Results: Few and Far Between

For all the noise the industry has made about wanting regulation, the years of lobbying have produced remarkably little. The repeated calls for clear rules have not translated into meaningful oversight, and the gap between rhetoric and reality keeps widening.

Consider what has actually changed:

• Promises of frameworks that never materialised
• Consultations that stalled without resolution
• Guidance that left the hardest questions unanswered

The pattern is consistent: talk is cheap, and action is scarce. Firms that publicly welcomed regulation have quietly benefited from its absence, while the public is left with the same unresolved risks that prompted the calls in the first place.

This is not to say nothing happened. But measured against the scale of the problem, the results are thin. The industry has had years to demonstrate good faith, and the record suggests that calling for regulation and accepting regulation are very different things.

What This History Means

The through-line is hard to miss: a technology arrives, alarm follows, and calls for regulation arrive almost as quickly. What rarely follows is meaningful regulation. Across successive waves of concern, the pattern has repeated — urgent warnings, promises of action, and then, more often than not, very little that changes how the industry actually operates.

That history should temper the current panic rather than inflame it. The present moment is not the first time these arguments have been made, and the fact that they feel unprecedented does not make them so. If past calls produced few results, there is little reason to assume this round will automatically break the pattern.

None of this means regulation is impossible or unnecessary. It means the burden of proof sits with those promising it. The relevant question is not whether the warnings are loud, but whether anything follows them.