Opaque AI Enforcement Hurts Compliant Labs More Than Bad Actors
Chollet's public warning signals that arbitrary regulatory strikes are becoming a competitive disadvantage for rule-following AI labs.
2. Opaque AI Enforcement Hurts Compliant Labs More Than Bad Actors
On June 4, 2026, François Chollet, creator of the ARC benchmark and a prominent voice in the AI research community, posted a public critique on X arguing that opaque and arbitrary regulatory enforcement actions are counterproductive for the entire AI industry, even from the perspective of regulation advocates. Chollet did not name a specific enforcement action, but the timing follows a period of intensifying regulatory activity across the EU AI Act implementation, the FTC's ongoing scrutiny of foundation model providers, and several state-level bills in California and Texas targeting AI deployment practices.
The strategic point Chollet is making is sharper than it first appears. Arbitrary enforcement, the kind that lacks clear criteria or advance notice, disproportionately burdens labs that invest in compliance infrastructure. OpenAI, Anthropic, and Google DeepMind have each built legal and policy teams, published model cards, and engaged regulators proactively. An enforcement regime that punishes visible actors on unclear grounds effectively subsidizes opaque competitors, including offshore developers operating outside U.S. or EU jurisdiction, who face no equivalent compliance cost. The result is a race to the bottom that regulation is supposed to prevent.
This fits a pattern worth tracking: practitioner frustration with regulatory process is shifting from private to public. When researchers with Chollet's credibility make this argument openly, it signals that the compliance-forward labs may be losing confidence that good-faith engagement pays off. Watch whether Anthropic or OpenAI begin coordinating more explicitly on regulatory process standards, not just policy outcomes. A push for procedural transparency requirements, clear safe-harbor definitions, and advance rulemaking notice could become the next lobbying priority for labs that currently have the most to lose from arbitrary strikes.
Source: @fchollet on X