The EU AI Act and Your Business: What You Need to Know, in Plain English

Learn about the EU AI Act and its impact on non-EU based companies.

Headquarters in Seattle. Obligations in Brussels. Why the EU AI Act reaches you, and what the dates actually are now.
This is not only an EU problem.
A US company with no European staff sells an AI resume-screening tool to a European retailer. EU job applicants are affected, so that company is a provider under the Act. Headquarters in Seattle. Obligations in Brussels.
Scope follows the output. If the result lands in the EU, the Act applies, and no European office is required for that to be true.
WHY THIS REACHES YOU
The Act regulates AI by use case and risk rather than by technology, which means the same tool can be minimal risk in one use and high risk in another. Obligations attach to roles. Providers build the system and deployers use it, and most organizations are deployers without having thought about what that means.
THE FOUR TIERS, IN PLAIN ENGLISH
Banned, regulated, disclose, and open. Where everyday business tools actually land, which is usually lower than people fear, and which uses quietly sit in the high-risk tier. Anything gating work, money, or services deserves a second look.
THE DATES THAT ACTUALLY MATTER
Several deadlines moved and several did not. Article 50 transparency duties applied from August 2, 2026. Standalone high-risk obligations under Annex III now land on December 2, 2027, moved back from August 2026. Embedded high-risk under Annex I moves to August 2, 2028. Regulatory sandboxes move to August 2027. The prohibited-practice bans and the GPAI duties are already in force, and the penalty ceilings and extraterritorial scope did not change at all.
WHAT THE PENALTIES LOOK LIKE
Up to 7 percent of global turnover or 35 million euros for prohibited practices, whichever is higher. Up to 3 percent or 15 million for provider and operator failures. Turnover means worldwide revenue, not EU revenue.
A WORLD TOUR
How the United States, Canada, and Brazil are handling the same problem, and why the pattern matters. Build to the EU standard and you will satisfy most of what is coming everywhere else. The reverse is not true.
WHO THIS IS FOR
General counsel, chief risk officers, compliance leads, and the executives who will be asked to sign off on AI systems somebody else selected. It is built for leaders rather than for lawyers, and no legal background is assumed.
WHAT YOU WALK AWAY WITH
An AI exposure checklist for your own systems. A plain-English map of the four tiers and where your tools likely sit. The current dates. Enough context to brief your leadership without a lawyer in the room.
This is general and directional. It is not legal advice and it is not a diagnosis of your specific exposure.
ABOUT THE PRESENTER
Adam Johnson is the founder of Rafer Consulting. He spent 25 years across Microsoft, Amazon, Meta, Dropbox, and Wheels Up building the data and decision infrastructure that preceded today's AI systems. He is an Executive Fellow at The Academy for AI Strategy and delivered the Architecture of Trust talk at Oxford.
One hour including live Q&A. Recorded for registered attendees.
Sign up here:
https://www.eventbrite.com/e/the-eu-ai-act-and-your-business-what-you-need-to-know-in-plain-english-tickets-2002355494185