News · December 9, 2023
EU negotiators agree on the AI Act, with phased obligations ahead
The European Parliament and EU member states reached a political deal on the AI Act, a risk-based law covering banned uses, high-risk systems and general-purpose models. Companies building or deploying AI for EU users now have a framework to plan against.
What happened
- According to the European Parliament, the deal bans uses such as social scoring, untargeted scraping of facial images and emotion recognition at work and in schools, and sets obligations for high-risk systems including a fundamental rights impact assessment.
- The Parliament said general-purpose AI models must provide technical documentation, comply with EU copyright law and publish summaries of training content, with stricter duties for high-impact models with systemic risk.
- Fines range from 35 million euros or 7% of global turnover down to 7.5 million euros or 1.5%, depending on the infringement and company size, and the deal promotes regulatory sandboxes to help SMEs, according to the Parliament.
- TechCrunch reported that the final text still had to be drafted and voted, and that obligations would phase in: prohibitions after six months, transparency and governance after 12 months and the rest after 24 months.
Why it matters
Most product teams will not build banned or high-risk systems, but many will use general-purpose models inside their products. The practical work is documentation: know which models you use, for what, with what data, and how you test and monitor them. That is good engineering practice regardless of the law.
Since the text was not final at this stage, we would not rebuild anything yet. Start an AI inventory, flag any features that touch hiring, credit, biometrics or other sensitive areas, and ask your model providers how they plan to support the new documentation requirements. For legal questions, talk to counsel.
Sources
- European Parliament: Artificial Intelligence Act: deal on comprehensive rules for trustworthy AI
- TechCrunch: EU lawmakers bag late night deal on 'global first' AI rules
This post is CMM’s own summary and commentary on publicly reported news. Facts are attributed to the sources linked above; please read them for full details. CMM is not affiliated with these publishers or with the companies mentioned, and trademarks belong to their owners. Prepared with AI assistance. Corrections or requests: [email protected]. See our editorial policy.
Want help with this?
Tell us what you need, and we will suggest the right team model on a short call.
Talk to us