The EU Now Fines for AI: Up to EUR 35M — What to Know If You Sell to Europe | inicIA
Direct answer
The EU AI Act is now enforceable: since August 2, 2026, the European Commission and its AI Office hold full enforcement powers, and fines are legally possible. The scheme: up to EUR 35 million or 7% of global annual turnover (whichever is higher) for prohibited practices; up to EUR 15 million or 3% for breaching other obligations (including transparency duties); up to EUR 7.5 million or 1% for giving misleading information to authorities. And it applies extraterritorially: a single European customer is enough to bring you in scope, wherever you are based. If your business uses AI and touches the European market, 2026 is the year to get your house in order.
What is already in force (and what is coming)
The AI Act rolls out in phases, and several dates have passed:
- August 2025: total ban on "unacceptable-risk" systems (social scoring, real-time remote biometric identification in public spaces).
- August 2026: enforcement powers active. The AI Office can request information and technical documentation, obtain direct access to models for evaluation, require corrective measures, and fine. The first enforcement letters, per September reports, target employment screening, credit scoring, and biometric identification systems.
- December 2, 2026: systems placed on the market before August must meet content-marking obligations (Article 50).
- 2027–2028: deadlines for high-risk systems (recruitment, HR) and new prohibited practices.
One detail many miss: chatbots and voice agents serving EU users must already identify themselves as AI at the start of the interaction, and AI-generated or manipulated content must carry machine-readable labels. This applies based on where your users are, not where your company sits.
Who it applies to (probably you too)
The AI Act reaches any company placing AI systems on the EU market or whose output is used within the EU, regardless of where it is incorporated. One European customer is enough. Typical cases for a business like yours:
- Your support chatbot serves visitors from Europe (it must identify itself as AI).
- You use an AI tool to screen resumes (high-risk system: the most scrutinized category).
- You generate marketing content with AI that circulates in Europe (labeling required).
- You resell or integrate a model from a European provider (value-chain cooperation duties, Article 25, now fineable).
For SMEs there is an adjustment: instead of "whichever is higher," the cap uses "whichever is lower" between the fixed figure and the percentage. It reduces exposure; it does not remove it.
The 5-step plan before year end
- Inventory. List every AI system you use or sell, including the ones embedded in products you already pay for (your CRM, your email tool, your chatbot).
- Prohibitions check. Confirm no tool falls under Article 5's banned practices (subliminal manipulation, social scoring, mass biometrics).
- Transparency. Check chatbots, intake bots, and AI-generated content against Article 50: clear identification and synthetic-content marking.
- Recruitment. If you use AI in hiring, that is your highest-risk front: ask vendors for their conformity plans and instructions for use, and update contracts.
- AI literacy. Document who was trained, on what, and when: the law requires literacy measures, and auditors ask for them.
Enforcement will be uneven (and that does not protect you)
The Commission's list of market surveillance authorities, updated September 7, 2026, still shows designations pending in a dozen Member States (including France, Germany, the Netherlands, Spain, and Sweden) and six countries with no listed authority at all. That is a gap in who enforces, not in what the law requires: your obligations exist regardless, and the Commission's AI Office acts centrally. Do not bet your business on "nobody is fining yet."
Related articles
- Privacy and data: what happens to your customers' information
- OpenAI Astra for Law: AI is already specializing by industry
- How to choose an AI vendor: 7 questions before signing
Frequently asked questions
I have a single European customer. Does it really apply to me?
Yes. The test is where the system's outputs are used, not where your company is based. One European customer brings you into scope.
What are the actual fines, in numbers?
Up to EUR 35M or 7% of global turnover for prohibited practices; up to EUR 15M or 3% for other obligations (including transparency); up to EUR 7.5M or 1% for misleading information to authorities. For SMEs, the cap uses the lower figure.
My chatbot only serves Spanish-speaking customers in Miami. Am I safe?
If no user is in the EU, the AI Act does not reach you through that channel. But if your site gets European traffic or you plan to expand, get it in order now: it is cheaper than fixing it later.
What happens on December 2, 2026?
AI systems placed on the market before August 2026 must meet Article 50's content-marking obligations: identifying AI-generated or AI-manipulated content.
Take the next step
If your business touches Europe — or plans to — and you use AI in support, sales, or hiring, get your compliance in order before someone asks for it. Talk to inicIA on WhatsApp: https://wa.me/17866120930
By inicIA
