8/11/2026, 1:05:08 PM · evaluation-safety

EU AI Act Enforcement Begins August 2, Requiring Chatbot Transparency Labels and Deepfake Disclosure

The European Commission's AI Office gained active enforcement powers on August 2, 2026, compelling transparency disclosures for chatbots and deepfakes while high-risk AI obligations for employment, credit, and education remain on a delayed track.

What Took Effect on August 2

<cite index="5-1">The European Union began enforcing key provisions of its Artificial Intelligence (AI) Act on August 2, requiring companies to disclose when people are interacting with AI and to label certain AI-generated or manipulated content.</cite>

<cite index="8-2,8-3">Article 50 of the EU AI Act requires providers to clearly inform people when they are directly interacting with an AI system unless that fact is obvious from the circumstances, and providers must also make covered AI-generated or manipulated content detectable through machine-readable markings.</cite>

<cite index="7-1">Deepfakes must be visibly disclosed, while AI-generated or manipulated text published to inform the public must be labelled when it has not undergone human review or editorial control.</cite> <cite index="7-7">People must also be informed when they are exposed to emotion-recognition or biometric-categorisation systems.</cite>

<cite index="2-15,2-16,2-17,2-18">AI that interacts directly with people — chatbots, virtual assistants, autonomous agents — must tell the user it is an AI the moment they first encounter it; fine print buried in terms of service will not suffice. The disclosure must be immediate, accessible, and clear. There are two exceptions: cases where it is obvious to anyone that they are dealing with an AI, and systems used for law-enforcement purposes.</cite>

Enforcement Architecture

<cite index="1-3">According to the European Commission's AI Act Service Desk, the AI Office and national authorities formally gained the power to enforce the rules on prohibited AI practices, transparency, and General-Purpose AI (GPAI) models from August 2, rather than simply monitoring compliance.</cite>

<cite index="11-7,11-8">On August 2, 2026, the Commission's enforcement powers in respect of GPAI model providers came into force. While the obligations of GPAI model providers came into force on August 2, 2025, providers were given an adjustment period of one year before the Commission could start exercising its supervision and enforcement powers against them.</cite>

<cite index="5-4">Enforcement is divided among the European Commission's AI Office, national regulators, and the European Data Protection Supervisor.</cite> <cite index="7-14,7-15">National market-surveillance authorities retain important responsibilities, while the EU AI Office has powers in relation to general-purpose models and certain systems built upon them. Cross-border cases may test how quickly these bodies exchange information and agree which authority should lead.</cite>

Penalty Structure

<cite index="35-10">EU AI Act fines run in three tiers under Article 99: up to €35 million or 7% of total worldwide annual turnover for prohibited AI practices, €15 million or 3% for most other breaches, and €7.5 million or 1% for supplying incorrect information.</cite> <cite index="36-8">Fines are calculated as the higher of the fixed euro amount or the revenue percentage for large organisations, and the lower of the two for small and medium-sized enterprises and startups — a critical proportionality mechanism.</cite>

The High-Risk Track: A Separate Timeline

The August 2 enforcement date does not mark a single compliance cliff for the AI Act as a whole. <cite index="23-3">The Digital Omnibus postpones the high-risk obligations for Annex III AI systems from August 2, 2026 to December 2, 2027, and the obligations for high-risk AI in regulated products under Annex I to August 2, 2028, but it leaves the Article 50 transparency rules and the Article 4 AI literacy duty exactly where they were.</cite>

<cite index="24-1">This legislative proposal, dubbed the 'AI Omnibus,' was adopted on November 19, 2025; a political agreement was reached on May 7, 2026, and it entered into force on July 27, 2026.</cite>

<cite index="18-4">Annex III covers the larger category for most enterprises: AI used in recruitment and employee management, creditworthiness assessment, biometric identification and categorization, education and vocational training access, migration and border control, law enforcement, and administration of justice.</cite> <cite index="21-6">That original August 2026 date was practically challenging, especially since the harmonised standards that providers need in order to demonstrate conformity were not going to be ready in time.</cite>

Code of Practice and Global Reach

<cite index="5-9">The Commission said more than 180 organizations have signed a voluntary Code of Practice designed to help companies apply the transparency rules for AI-generated content.</cite> <cite index="14-12">For signatories, the Commission will focus enforcement activities on monitoring adherence to the Code and may take commitments to the Code into account as mitigating factors when fixing the amount of fines; non-signatories are expected to demonstrate compliance via other adequate means.</cite>

<cite index="8-11,8-12">The practical consequence extends beyond European technology companies. American developers, platforms, retailers, and service providers offering AI products in the European market may need to adjust user notices and content provenance practices.</cite> <cite index="8-12">Generative AI systems placed on the market before August 2 may receive until December 2026 to satisfy the machine-readable marking requirement.</cite>

Cross-references

Sources

  1. [1]
    Chatbots Must Now Confess They're AI and Deepfakes Need Labels Under New EU Rules | IBTimes UK
  2. [2]
    EU AI Act Aug 2: Chatbot & Deepfake Transparency | Pebblous
  3. [3]
    Commission starts enforcing AI Act rules and new transparency requirements on 2 August | Shaping Europe’s digital future
  4. [4]
    The EU AI Act’s Transparency Rules: A Practical Guide to Article 50 | EU Artificial Intelligence Act
  5. [5]
    EU Begins Enforcing AI Act as Transparency Rules Take Effect
  6. [6]
    AI Act | Shaping Europe's digital future - European Union
  7. [7]
    EU’s AI transparency rules bring chatbot and deepfake labels into force - https://eutoday.net
  8. [8]
    EU AI Transparency Rules Take Effect With New Labels for Bots and Deepfakes
  9. [9]
    The Era of Enforcement: EU AI Act Redraws the Global Map for Artificial Intelligence
  10. [10]
    Enforcement of Chapter V under the EU AI Act | EU Artificial Intelligence Act
  11. [11]
    Generally Speaking: Does Your Company Have EU AI Act Compliance Obligations as a General-Purpose AI Model Provider? | Advisories | Arnold & Porter
  12. [12]
    EU AI Obligations for GPAI Providers: Compliance, Enforcement & Deadlines (2025–2027) - MediaLaws
  13. [13]
    An Introduction to the Code of Practice for General-Purpose AI | EU Artificial Intelligence Act
  14. [14]
    EU AI Act: What Actually Applies on August 2, 2026
  15. [15]
    Implementation Timeline | EU Artificial Intelligence Act
  16. [16]
    The EU AI Act Newsletter #82: GPAI Code of Practice Goes Live
  17. [17]
    EU AI Act Digital Omnibus: The New High-Risk AI Deadlines After Council Approval | Secure Privacy Blog
  18. [18]
    The Digital AI Omnibus: Proposed deferral of high risk AI obligations under the AI Act (update) - DLA Piper GENIE
  19. [19]
    EU AI Act Omnibus Agreement — Postponed High-Risk Deadlines and Other Key Changes - Gibson Dunn
  20. [20]
    EU AI Act unpacked #34: The final Digital Omnibus on AI - Key amendments to the AI Act and their impact on businesses active in the EU | Freshfields
  21. [21]
    EU Nears Approval of Agreement to Delay Rules for AI Use in Employment Decisions - Ogletree
  22. [22]
    The Digital Omnibus and the postponement of high-risk obligations to December 2027: what changes and what still applies
  23. [23]
    EU AI Act omnibus: the new high-risk deadlines explained
  24. [24]
    The EU AI Act and assessment: December 2027 is not a snooze button
  25. [25]
    AI Act rules on high-risk AI delayed as AI Digital Omnibus agreed - Winston Taylor
  26. [26]
    EU AI Act Fines & Penalties: The Complete Guide
  27. [27]
    Article 101: Fines for Providers of General-Purpose AI Models | EU Artificial Intelligence Act
  28. [28]
    EU AI Act 2026: Penalties, Risk Tiers & New Deadlines
  29. [29]
    Penalties of the EU AI Act: The High Cost of Non-Compliance
  30. [30]
    Enforcement / fines in the European Union - AI Laws of the World
  31. [31]
    Article 99: Penalties | EU Artificial Intelligence Act
  32. [32]
    EU AI Act Penalties and Enforcement
  33. [33]
    EU AI Act Fines and Penalties: How They Work (2026) — Insights | easyAI
  34. [34]
    Legalithm
  35. [35]
    EU AI Act Fines and Penalties: What Non-Compliance Will Cost You | Matproof Blog