America and Europe Have Taken Different Routes to Control AI — The Results Are Stark

America is deregulating AI while Europe enforces its AI Act. Fortune says the results are stark. Here is the complete comparison of what each approach has produced and which is actually winning.

## Two Philosophies, Two Outcomes, One Technology

When the European Union's AI Act entered its full application phase in 2026, it represented the world's most comprehensive attempt to regulate artificial intelligence through risk-based categorization — high-risk AI systems subject to specific conformity assessments, prohibited AI applications banned outright, and transparency requirements imposed across the full range of AI deployments. The regulatory approach reflected a specific European philosophy about technology: that powerful new capabilities should be governed before they create harms that retroactive regulation cannot adequately address.

The United States, under the Trump administration, moved in the opposite direction. President Trump's December 2025 executive order on AI removed specific Biden-era AI safety requirements, reduced regulatory oversight mechanisms, and explicitly framed American AI development as a competitive race against China whose outcome depends on reducing not increasing the regulatory friction that governance creates. Fortune magazine's April 2026 characterization of the two approaches — "America and Europe have taken different routes on trying to 'control AI.' The results are stark" — captures the specific observable difference that one year of divergent regulatory environments has produced.

## What American Deregulation Has Produced

On the American side, the results include explosive revenue growth at the companies whose development the deregulatory environment has freed from specific compliance burdens. OpenAI's annualized revenue reached $25 billion in 2026, with 900 million weekly active users. Anthropic reached $19 billion in annualized revenue. Microsoft, Google, Amazon, and Meta collectively invested hundreds of billions in AI infrastructure, with the specific Stargate initiative representing $500 billion in planned infrastructure investment.

The competitive position of American AI companies in global markets has been maintained — though complicated by Chinese companies' specific open-source strategy, which has achieved the particular result of making Chinese AI models widely adopted by American and European startups regardless of government policy preference. DeepSeek's specific models, and the subsequent Chinese open-source wave, have created a specific competitive dynamic that neither deregulation nor regulation fully addresses.

The safety concerns that American AI governance removed from mandatory assessment have produced specific incidents whose accumulation is creating retroactive pressure. MIT Technology Review's specific flagging of the pending OpenAI trial related to an AI system and teenager suicide represents the particular legal and ethical dimension whose resolution the American regulatory approach deferred rather than prevented. The specific cases that will reach trial in 2026 — AI companies potentially liable for what their chatbots encourage users to do — are the retroactive governance that deregulation's critics predicted would follow.

## What European Regulation Has Produced

Europe's AI Act implementation has produced the specific compliance burden that its critics predicted: higher implementation costs for companies operating in the European market, specific delays in deploying systems that require conformity assessments, and the particular uncertainty around what constitutes a "high-risk AI system" whose legal definition the Act's provisions are still being clarified through regulatory guidance and early case law.

But the specific European companies — and American companies that adapted their European deployments to AI Act requirements — have produced a different kind of product: one with documentation of training data sources, specific explainability requirements for consequential decisions, and the particular bias assessment whose absence from American equivalents is increasingly visible as the first wave of AI-driven consequential decisions (loan approvals, hiring, medical diagnostics) encounters specific retrospective legal challenge.

The specific advantage that Europe's approach may produce over the medium term involves trust. The particular consumer confidence that comes from knowing that AI systems making consequential decisions have been specifically assessed for their risk level creates the specific adoption environment in sensitive sectors — healthcare, financial services, criminal justice — whose governance requirements make compliance documentation a commercial advantage rather than merely a regulatory cost.

## The Race and Its Actual Outcome

The Fortune analysis suggests neither approach is clearly winning on the specific metrics that matter most. American AI companies lead in revenue, user numbers, and raw capability benchmarks. European AI governance leads in accountability documentation and the specific institutional readiness for what MIT Technology Review describes as "AI's upcoming legal battles" — the messier, more consequential litigation that determines whether AI companies are liable for the specific harms their systems produce.

What's becoming clear is that the specific framing of regulation-as-impediment-to-competition was always partially incorrect. The EU's data privacy regulations (GDPR), which American critics called a competitive disadvantage in 2016, are now globally adopted standards that created specific American competitive advantages in markets where data protection is a genuine consumer priority. The AI Act may follow a similar trajectory.