< Back to all clusters
[TECHNOLOGY] · 15 sources

European Commission Enforces AI Act Transparency Rules, Fines Up to €35 Million

On 2 August 2026 the European Union’s AI Act brought its transparency obligations into force. Companies must now inform users when they are interacting with an AI system and must label any AI‑generated content – including images, video, audio and deepfakes – with machine‑readable watermarks or icons. The deadline for applying these labels to synthetic media is 2 December 2026. Non‑compliance can attract fines of up to €35 million for serious breaches and up to €15 million (or 3 % of global turnover) for lesser violations. The new European AI Office gains authority to assess large AI models and demand access from providers, while national regulators will enforce the rules on smaller models. Providers are also required to carry out risk assessments and put mitigation measures in place. The regulations affect businesses across the EU, with examples cited from Spain, Portugal, the Netherlands and other member states.

Entities: AI Office · Anthropic · European AI Office · European Commission · Google · OpenAI

Claims

What the coverage asserts, and how well corroborated each claim is across sources.

  • [● 9 SOURCES] AI‑generated content must be labeled by 2 December 2026. (EU AI Act)
  • [○ 1 SOURCE] From 2 December, AI‑generated text, images, video and audio must include watermarks indicating manipulation. (Article 3 (Portuguese))
  • [○ 1 SOURCE] OpenAI, Google and Anthropic are subject to the EU AI Act’s transparency and labeling obligations. (Article 3 (Portuguese))
  • [● 4 SOURCES] National regulators will enforce the AI Act rules on smaller AI models. (EU AI Act)
  • [● 5 SOURCES] AI‑generated content must be clearly identified, with deepfakes labeled by icons or warnings. (EU AI Act)
  • [● 4 SOURCES] The European AI Office can evaluate large AI models and require provider access. (EU AI Act)
  • [● 3 SOURCES] The EU AI Act’s new transparency obligations entered into force on a Sunday. (All three articles)
  • [● 5 SOURCES] Providers of AI models must conduct risk assessments and adopt mitigation measures. (EU AI Act)
  • [● 4 SOURCES] Companies that violate the transparency rules can be fined up to €35 million for serious infringements. (EU AI Act)
  • [● 3 SOURCES] Fines of up to €15 million or 3 % of global turnover apply for less severe violations. (EU AI Act)
  • [● 9 SOURCES] The EU AI Act's transparency obligations entered into force on 2 August 2026. (EU AI Act)

Sources

about 5 hours ago
about 3 hours ago