← Insights / Compliance

Those EU AI Act Fines You Have Read About Did Not Happen

Reports of the first EU AI Act fines are spreading, including against hiring AI. Two facts from the EU's own sources show they cannot be true. Here is how to check one.

Compliance 22 September 2026 2 min read

Search for EU AI Act enforcement this month and you will find confident reports of the first fines: tens of millions of euros against a hiring-software company, a credit scorer, a retail chain using emotion recognition. Some are dated as early as March 2026. They read like news. Two facts from the EU's own sources show they cannot be.


Fact one: enforcement started on 2 August 2026

The European Commission announced that the AI Office, together with national authorities, began enforcing the AI Act on 2 August 2026. A fine "issued in March 2026" by the AI Office pre-dates the power to issue it. Any story with that date is wrong on its face.

Fact two: the hiring-AI rules do not apply yet

The most shared of these stories describe a company fined for using AI in recruitment without the documentation, conformity assessment or human oversight the Act requires of high-risk systems. Those are real obligations — but the Digital Omnibus, Regulation (EU) 2026/1744, deferred them. For AI used in recruitment and the other Annex III uses, they apply from 2 December 2027. Nobody can be fined today for breaking a rule that does not yet apply.

There is also a question of who. The AI Office supervises providers of general-purpose AI models. A company using a hiring tool would answer to a national authority, not the AI Office.


Why this matters to a small firm

Stories like these do real damage in both directions. Some firms panic and buy expensive programmes they do not need yet. Others eventually learn the stories were invented and decide the whole Act is hype. Both are wrong.

What is true today:

Some parts of the Act already apply. The prohibited practices — including AI that reads people's emotions in the workplace — and Article 4 on AI literacy are in force now.

The high-risk rules have a fixed date. If you use AI to screen candidates, 2 December 2027 is the date your deployer duties begin. That is time to prepare properly, not a reason to ignore it.

Enforcement stories are checkable. A real fine comes with an authority's name and its own announcement. If the only source is a blog or a post, treat it as unconfirmed.


A simple test before you share one

Ask three questions. Which authority issued it, and where is its announcement? Did the rule it describes apply on the date given? Is that authority the one responsible for that kind of business? A story that fails any one of them is not news.


Know what actually applies to you

The free AI risk checker at opsintel.io tells you which parts of the EU AI Act reach your business today and which start later. Ops Intel writes AI compliance frameworks for small and medium-sized businesses, built on the law as it currently stands.

Follow us in Google

See Ops Intel first when AI rules change

One click tells Google you want our compliance briefings near the top of your results — and marks us as a preferred source inside Google's AI answers. It applies to your Google account only, and you can undo it any time.

What to do about it

The news is what changed. A framework is what you do about it.

Ops Intel writes AI compliance frameworks for small and medium businesses worldwide. Before you spend anything, read a real one — the whole pack, produced by the same system that will write yours.

Call Now See prices