# Big Tech Aims to Lock In AI Dominance Through Regulatory Capture
Sam Altman, the chief executive of OpenAI, has emerged as the public face of a corporate strategy that wraps monopolistic control in the language of safety and responsibility. His push for heavy government regulation of artificial intelligence development contains a calculated trap: stricter rules benefit entrenched players with the resources to comply while crushing smaller competitors and open-source developers who lack massive balance sheets.
The proposal functions as regulatory capture disguised as consumer protection. Large AI companies like OpenAI, Google, and Anthropic possess the capital, legal teams, and compliance infrastructure to navigate onerous federal requirements. Startups and independent researchers cannot. The effect is predetermined: consolidation at the top, barriers to entry everywhere else, and a government-blessed oligopoly controlling one of the most transformative technologies of the era.
Altman's specific advocacy centers on strict licensing requirements for models above certain capability thresholds. On the surface, this targets genuinely powerful systems that pose real risks. But the thresholds themselves become negotiable battlegrounds. Companies influence regulators. The definitions shift to encompass competitors' work while exempting their own. A company worth hundreds of billions shapes the rules to protect its position.
The alternative framework involves tiered regulation, not blanket restriction. High-capability models with genuine dual-use risks warrant serious oversight. Governments should require transparency, testing, and auditing before deployment at scale. But this does not require killing open-source AI development, which has produced some of the field's most innovative work and checked corporate power.
Open-source models democratize access to AI capability. They enable researchers at universities, nonprofits, and small companies to build applications, test safety measures, and innovate without seeking permission from Silicon Valley gatekeepers. They create competition that forces dominant players to improve products and reduce prices. A thriving open-source ecosystem is not a bug to be regulated away. It is essential infrastructure for a competitive market.
The two-tier approach preserves both goals. Governments establish strict requirements for closed, proprietary models reaching high capability levels. These requirements include safety testing, bias audits, and restrictions on dangerous applications. Meanwhile, open-source models below those thresholds operate with lighter oversight, enabling continued innovation and competition. The threshold itself reflects genuine technical risk, not corporate convenience.
This distinction matters because it accepts the legitimate case for AI safety regulation while rejecting the illegitimate case for corporate monopoly. Real risks exist. Models trained on billions of parameters with unclear failure modes warrant scrutiny. But the answer is not to hand OpenAI, Google, and Anthropic permanent control over which AI systems exist and who can build them.
Altman frames his position as necessary caution in a dangerous field. That framing obscures the play underneath. Every regulation shapes incentives. Every barrier to entry benefits incumbents. Every licensing requirement concentrates power. When a company worth $80 billion advocates for heavy government involvement in its industry, the motive deserves skepticism. Regulation designed by those it regulates serves those who designed it.