New EU AI Act transparency obligations took effect on 2 August 2026, requiring providers to clearly label AI-generated or manipulated content — including images, audio and video resembling real people (“deepfakes”), emotion-recognition and biometric-categorisation tools, and AI-generated text on public-interest matters lacking human review. Users must also be informed when they are interacting with an AI system such as a chatbot, agent or avatar. Companies can face fines of up to €15 million or 3% of global annual turnover, whichever is higher; EU institutions can be fined up to €750,000, the European Commission said.
On 11 August, Euronews reported that Anthropic will invisibly watermark all of Claude’s text and file output worldwide to meet the new rules. Anthropic said: “Claude models launched on or after August 2, 2026 support marking at launch. We’re also working to add marking support to Claude models released before that date.” Text carries “an imperceptible watermark directly into the text itself”; supported files (.svg, .png, .jpg) get signed provenance metadata. Marking covers the Claude consumer app, Claude Platform (API), Claude Code, Claude Cowork, Claude Tag, and versions accessed through AWS, Google Cloud and Microsoft Foundry.
