US AI deregulation will not save you from EU rules

In December 2025, the US and the EU moved in opposite directions on AI in the same week. Washington pushed for speed. Brussels pushed for compliance.

If you sell to any customer outside the US, US deregulation does not free you from European rules. The Brussels effect makes the strictest market the global standard, because two compliance stacks cost too much. Build your AI systems to EU standards for data privacy and transparency now. That is the cheaper path, and it gives you a scalable, sellable asset.

What did the US and the EU decide in December 2025?

The US chose one permissive federal rule. The EU chose more enforcement.

In the US, President Trump announced an executive order to set a single national "One Rulebook" for AI. The order intends to preempt the growing patchwork of state-level AI rules in places like California and Colorado. The order he signed on December 11, 2025 tells the Attorney General to set up a task force that challenges state AI laws, and it names a Colorado law as an example. For US businesses that build AI tools, this means one permissive federal standard instead of 50 different state laws.

In the same week, the European Commission opened an antitrust investigation into Google. The question is whether Google uses content from publishers and YouTube creators for its AI services without adequate compensation. The EU is doubling down on compliance while the US deregulates.

Why US deregulation creates a splinternet problem

The split creates a splinternet problem for business owners. A splinternet is an internet where the same tools follow different rules in different regions. If you operate globally, your AI tools may soon work differently in the EU than in the US, because of data usage restrictions.

You might think US deregulation gives you freedom. Let's be honest: if you have any international customers, the Brussels effect probably applies to you.

Why the strictest market sets the rules

The Brussels effect, a term Columbia Law professor Anu Bradford coined in 2012, is when the strictest market, usually the EU, sets the global standard in practice. Companies follow the strictest rules everywhere because it is too expensive to maintain two different compliance stacks. We saw this with GDPR. A company that served European customers did not build a second, looser data system for everyone else.

AI rules follow the same economics. Two versions of every workflow, every data store and every privacy notice add to your build and your maintenance costs.

How to build your AI strategy for both markets

My advice: do not build your AI strategy around the promise of US deregulation. Build to the stricter European standards now.

1. Check where your customers are. If even one market is in the EU, assume EU rules apply to your AI systems.
2. Use the stricter standard as your baseline. Design for EU data privacy and transparency rules first. One compliance stack is cheaper than two.
3. Document your data flows. Write down which data each AI tool uses and where that data comes from. EU regulators are already asking where AI training data comes from, as the Google probe shows.
4. Treat compliance as part of the asset. A system built to the strictest rules is easier to scale and easier to sell.

For a practical starting point, use the EU AI Act Compliance Checklist. If you run AI agents for clients, also read who pays when your AI agent breaks a client's data.

The bottom line

US deregulation is about speed. EU regulation sets the standard for anyone who sells across borders. Building to EU standards now is future-proofing, not only compliance.

Build a calmer business,Yuri

Frequently asked questions

What is Trump's One Rulebook executive order on AI?

In December 2025, President Trump announced an executive order to set a single national "One Rulebook" for AI in the US. The order intends to preempt the state-level AI rules that were growing in places like California and Colorado. For US businesses, it points to one permissive federal standard instead of 50 different state laws.

What is the Brussels effect in AI regulation?

The Brussels effect is when the strictest market, usually the EU, sets the global standard in practice. It happens because it is too expensive for companies to maintain two different compliance stacks, one for each market. GDPR is the best-known example. AI rules for data privacy and transparency can follow the same pattern.

Does US AI deregulation mean my business can ignore EU AI rules?

No. US deregulation in December 2025 aims at speed, but the EU went the other way and opened an antitrust probe into Google over publisher data. If your business has any international customers, the Brussels effect probably applies to you. Build to the stricter European standards for data privacy and transparency.

Should a small business build its AI systems to US or EU standards?

Build to EU standards if you have any customers outside the US. Two compliance stacks cost more than one, and the strictest market usually sets the global standard, as GDPR did. A system built to EU rules for data privacy and transparency is easier to scale and easier to sell. It is future-proofing, not only compliance.