All insights · Guide
The EU AI Act for SMEs: what you actually need to do
Most of the noise about the EU AI Act is about big tech. If you run a 20 or 50-person business that uses ChatGPT, Copilot or the AI built into your software, your list is much shorter. Here it is, as the law stands in November 2026.
Start with your role
The AI Act separates providers, who build or sell AI systems, from deployers, who use them in their business. If your team uses tools made by someone else, you are a deployer. Most of the heavy obligations sit with providers.
What already applies to you
- A few uses are banned. A short list of practices has been prohibited since 2 February 2025. Few small businesses come anywhere near them. A new ban on tools that generate non-consensual intimate images applies from 2 December 2026 (FPF timeline).
- Your team needs basic AI training. Article 4 asks you to take measures that support AI literacy among the people who use AI for you. Since the July 2026 amendments, you no longer have to guarantee a set level for each person. In practice, keep a record of the training and awareness you provide (Lewis Silkin).
- People must know when AI is involved. The transparency rules apply from 2 August 2026. A chatbot your customers talk to must make clear it is an AI. If you publish AI-made images, audio or video that look real, you must say so. AI-written text published to inform the public needs a label too, unless a person reviewed it and someone takes editorial responsibility (Article 50).
- Fines are scaled for small firms. For SMEs, each fine is capped at whichever is lower: the fixed amount or the percentage of turnover (Article 99).
What can wait, but not forever
The strict "high-risk" rules cover uses such as screening job candidates or evaluating staff. They were due in August 2026 and now apply from 2 December 2027 (Lewis Silkin). If you use AI in hiring or HR decisions, use that time to understand what will be expected of you as a user.
A plan for one afternoon
- List the AI tools your team uses. Ask people directly, and check your software subscriptions for AI features.
- Flag anything that touches hiring, staff evaluation or decisions about customers.
- Write a one-page AI policy: which tools are allowed, what data never goes into them, who to ask.
- Run a short session with your team and keep a note of who attended.
- Check that any customer-facing chatbot and any AI content you publish is clearly labelled.
Sources
- FPF, AI Act timeline after the Omnibus
- Lewis Silkin, the Digital Omnibus on AI enters into force
- European Commission AI Act Service Desk, Article 50
- European Commission AI Act Service Desk, Article 99
Related reading
Want to talk this through for your business?
Let's talk