Europe’s AI Act now requires disclosure: chatbots, deepfakes and biometrics

Since August 2, 2026, the European Union’s Artificial Intelligence Act activates its transparency obligations and its sanctioning regime. The rules of Article 50 of Regulation (EU) 2024/1689, the first global legal framework for AI, are now enforceable for providers and deployers.

What changes

• Users must be immediately notified when they interact with a chatbot, without ambiguity.
• Content generated or manipulated by AI —deepfakes, texts, images— must be labeled as such.
• The analysis of biometric data must be reported immediately to the affected person.

What is postponed

The most complex requirements for systems considered high risk have been postponed to 2027 and 2028, following the regulatory changes known as the Omnibus package. The European Commission maintains the AI Pact, a voluntary path for companies to move forward before the mandatory deadlines.

Why it matters

The European law is the global reference: although the Dominican Republic is not an EU member, local companies operating in Europe or developing AI for that market will have to comply, and its rules set the trend for regulators in the region.

The key figure: August 2, 2026: transparency and sanctions begin; high-risk requirements arrive in 2027-2028.