AI regulation is moving fast, and not in a straight line. In the past year, the EU delayed its biggest AI rules, Colorado repealed and replaced its landmark AI law, new state laws took effect in Texas and Illinois, and the federal government launched an effort to challenge state AI laws in court. If you’re a business owner trying to figure out what actually applies to you, it’s confusing.
Here’s the practical picture as of October 2026: which rules matter for ordinary businesses, the key dates, and what to do now. Use the checker to see which ones are likely relevant to you.
Key takeaways
- There’s still no single federal AI law in the U.S. Existing laws on discrimination, privacy and consumer protection already apply to AI.
- The riskiest uses are decisions about people: hiring, lending, housing, insurance. That’s where most new rules focus.
- Illinois and Texas laws are in effect now. California and Colorado obligations arrive January 1, 2027.
- The EU AI Act’s transparency rules apply now. Its high-risk rules were delayed to December 2027.
- Good practice covers most of it: know where you use AI, keep humans in key decisions, be transparent, and test for bias.
Which rules might apply to you?
Interactive
AI rules checker
Key dates at a glance
| Date | What happens |
|---|---|
| Since July 2023 | NYC Local Law 144: bias audits and notices for automated hiring tools |
| Since Feb 2025 | EU AI Act: banned practices and AI-literacy duty apply |
| Jan 1, 2026 | Texas TRAIGA and Illinois HB 3773 take effect; California automated decision-making rules take effect (with a 2027 compliance date) |
| Aug 2, 2026 | EU AI Act transparency duties for chatbots and AI-generated content apply |
| Jan 1, 2027 | California automated decision-making compliance deadline for existing uses; Colorado SB 26-189 scheduled to take effect |
| Dec 2, 2027 | EU AI Act high-risk obligations (employment, credit, education and more), delayed from August 2026 |
| Aug 2, 2028 | EU AI Act rules for AI embedded in regulated products |
The federal picture
Congress hasn’t passed a comprehensive AI law, and attempts to block state AI laws through federal legislation have so far failed. In December 2025, an executive order directed the Department of Justice to form an AI Litigation Task Force to challenge state AI laws. In April 2026, a federal court paused enforcement of Colorado’s original AI Act after a legal challenge the DOJ joined. Colorado has since replaced that law.
The practical advice from most legal observers: keep complying with state laws as written. An executive order can’t overturn a state statute. Only Congress or the courts can, and that process takes time.
What to do now: a practical checklist
Checklist
AI compliance basics
Frequently asked questions
Does the EU AI Act apply to U.S. businesses?
It can. The Act applies to organizations that place AI systems on the EU market or whose AI output is used in the EU, regardless of where they’re based. If you have EU customers or employees, check with counsel.
If a vendor’s AI tool discriminates, is it the vendor’s problem?
Not only theirs. Employers and businesses are generally responsible for decisions they make, even when a vendor’s tool is involved. Ask vendors about bias testing, keep humans in the loop, and review contracts for responsibilities and indemnities.
We only use AI for drafting emails and documents. Do these laws affect us?
Mostly not. Most new rules target consequential decisions about people and AI that interacts with or is shown to the public. For internal productivity, focus on data protection, accuracy and an AI use policy.
Free template
Put your AI rules in writing
Our editable AI acceptable use policy covers approved tools, data rules, human review and training: the foundation regulators expect to see.
