EU enforces new AI transparency rules for companies
The European Union’s AI Act entered a new phase on 2 August 2026, imposing transparency obligations on companies that use artificial‑intelligence tools. Under the rules, AI‑generated or AI‑manipulated images, video, audio and text must be clearly labelled – for example with watermarks, metadata or on‑screen notices – so that users can distinguish synthetic content from real material. The regulation also requires that users be informed when they are interacting with a chatbot or other AI system, unless the AI nature is already obvious.
Exemptions apply to routine post‑production edits, internal documents, marketing copy that is reviewed by a human, and artistic or satirical works. Companies must provide AI‑literacy training for staff, update internal procedures and contracts, and retain technical information that enables the origin of AI‑generated content to be identified. The EU’s AI Office now has expanded supervisory powers and can impose fines for non‑compliance. The measures aim to curb deep‑fakes, disinformation and high‑risk AI applications while preserving innovation across the bloc.
Entities: AI Act · Andrea Moletto · Calero · Companies · European Commission · European Union · Frank Fischer · Google · Juan Pablo Figueroa · OpenAI
Claims
What the coverage asserts, and how well corroborated each claim is across sources.
- [○ 1 SOURCE] The AI Act classifies AI systems into risk levels: minimal, limited, high and unacceptable. (07c1bd23-cdfd-4647-8c15-120754088a44)
- [○ 1 SOURCE] The European Commission’s AI Office now has powers to supervise generative‑AI models and impose fines for non‑compliance. (59f80163-37fc-4694-a95f-21aba6773094)
- [○ 1 SOURCE] From 2 December 2024 AI systems must incorporate watermarks for manipulated text, images, video and audio that could mislead the public. (59f80163-37fc-4694-a95f-21aba6773094)
- [● 2 SOURCES] Companies must ensure AI literacy training for employees who use AI tools. (2843a169-c6b1-404c-b7d6-23a0448e8155)
- [○ 1 SOURCE] Chatbots must inform users they are interacting with AI unless the interaction is already evident. (3c209258-8b2c-4038-a488-738a4257f16d)
- [○ 1 SOURCE] The AI Act prohibits high‑risk practices such as subliminal manipulation, targeting vulnerable groups and social‑scoring. (2843a169-c6b1-404c-b7d6-23a0448e8155)
- [○ 1 SOURCE] The first part of the EU AI Act entered into force on 2 February 2025. (2843a169-c6b1-404c-b7d6-23a0448e8155)
- [● 2 SOURCES] From 2 August 2024 a new phase of the AI Act requires AI‑generated content to be clearly labelled or watermarked. (3c209258-8b2c-4038-a488-738a4257f16d)