When Does the EU AI Act Apply to Non-EU Businesses?

It applies from the point at which you do something market-facing in the EU, not from the point at which you establish a presence there. The triggers are placing an AI system on the EU market or putting it into service in the Union, deploying one from an establishment in the EU, or, where you sit outside the EU entirely, producing output with an AI system that is then used in the Union. Selling to a European customer, embedding an AI feature in a product available in the EU, or generating output that your European customers rely on can each be enough on its own.

The timing is phased, so the more useful question is which obligations apply now. Prohibitions on unacceptable practices and the AI literacy provisions applied from 2nd February 2025, and obligations for providers of general-purpose AI models from 2nd August 2025. Enforcement by the Commission's AI Office and national authorities began on 2 August 2026, alongside the transparency rules covering chatbots, deepfakes and machine-readable marking of AI-generated content. The Digital Omnibus on AI, in force since 27th July 2026, moved the obligations for standalone high-risk systems listed in Annex III to 2nd December 2027 and those for AI embedded in regulated products to 2nd August 2028. New prohibitions on AI systems generating non-consensual sexually explicit content and child sexual abuse material apply from 2nd December 2026.

For most US and UK SaaS businesses selling into Europe, the date that matters has already passed. If your product includes a chatbot, generates text, or produces synthetic images, audio or video, the transparency obligations apply to you now. The high-risk deadlines give you more room, but only once you have established that none of your systems fall within Annex III, which is a determination worth documenting rather than assuming, particularly if your product touches employment, education, creditworthiness or access to essential services.