AI Literacy / Glossary

EU AI Act

The EU AI Act is the European Union's law on artificial intelligence. It sorts AI uses by risk, bans a few harmful practices, sets strict rules for high-risk uses and requires transparency for chatbots and deepfakes.

The law entered into force in August 2024 and applies in stages. Bans on certain practices, such as social scoring and manipulative AI that exploits vulnerabilities, have applied since 2 February 2025. So has a duty on organisations that provide or use AI to take measures for their staff's AI literacy; since July 2026 it is worded as supporting AI literacy, with no guaranteed level.

Uses are grouped by risk: unacceptable (banned), high risk (for example AI used in hiring, education, credit or essential services, which needs testing, documentation and human oversight), limited risk (transparency, such as telling people they are talking to a chatbot) and minimal risk (most everyday AI).

Transparency duties, such as chatbot disclosure and deepfake labels, apply from 2 August 2026. A 2026 EU simplification package (the "digital omnibus") moved most high-risk rules to December 2027 or August 2028, and each country names its own supervisory authorities. For people in the EU, the practical point is simple: you should be told when you are dealing with AI, and important AI decisions about you need human oversight.