Background
<cite index="19-2">The European Union's Artificial Intelligence Act (Regulation 2024/1689) is the world's first comprehensive AI law.</cite> <cite index="19-3">It classifies AI systems into four risk tiers — unacceptable, high, limited, and minimal — with fines up to €35 million or 7% of global annual turnover under Article 99.</cite> The Act has been rolling into effect in phased stages since it entered into force in August 2024, with different obligations applying at different dates.
What Took Effect on August 2, 2026
<cite index="12-1,12-2">Article 50 of the AI Act applies from 2 August 2026. It sets out transparency obligations for providers and deployers of certain AI systems, including generative and interactive AI systems and deepfakes.</cite>
The European Commission's own digital strategy page confirmed the scope on the date of enforcement. <cite index="5-4">Under the new rules, chatbots and other interactive AI systems must tell users they are dealing with AI, not a human.</cite> <cite index="5-5,5-6">Deepfakes — images, videos, or audio that have been edited or generated using AI — must be labelled, and AI-generated or altered content must also carry machine-readable marks so it can be detected more easily.</cite>
<cite index="3-3">Providers of interactive AI systems must design them to inform users that they are interacting with a machine unless this is obvious from the circumstances.</cite> <cite index="3-5,3-6">Separate duties fall on organisations deploying the technology: 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="3-7">People must also be informed when they are exposed to emotion-recognition or biometric-categorisation systems.</cite>
<cite index="6-5">The requirements apply to both European companies and foreign products aimed at users in the European Union.</cite>
Enforcement Mechanism
<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 2 August.</cite> <cite index="23-7">From that date, the Commission can move from persuasion to compulsion: requesting documentation, evaluating models directly, ordering corrective measures, restricting or withdrawing models from the EU market, and imposing fines.</cite>
<cite index="21-4">The Article 99 penalty framework puts GPAI violations and Article 50 transparency violations in the same maximum tier: fines up to the greater of €15 million or 3% of total worldwide annual turnover for the preceding financial year.</cite> <cite index="21-11">Article 50 enforcement sits with national market surveillance authorities from August 2, 2026.</cite>
A limited transitional carve-out exists for legacy systems. <cite index="10-3,10-4">Chatbot disclosure, machine-readable marking of AI-generated content, and deepfake labeling were not postponed. The only carve-out is that pre-existing systems get until December 2, 2026 on the machine-readable watermarking duty.</cite>
Industry Response and Voluntary Measures
<cite index="4-5">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="20-8,20-9">Signing the voluntary Code of Practice does not buy immunity from fines, but the AI Office has said it will weigh a provider's good-faith adherence to the Code when assessing penalties.</cite>
Scope Caveats
<cite index="3-8,3-9,3-10">The start date should not be described as the full AI Act entering into force. Brussels has extended the timetable for high-risk systems under its AI Omnibus, with stand-alone high-risk systems now due to face the relevant obligations from December 2027.</cite> <cite index="3-12">Transparency duties for consumer-facing and generative systems are becoming applicable while some of the most demanding rules for employment, education, critical infrastructure, and regulated products remain in transition.</cite>
<cite index="7-13">Industry groups have raised a concern that late guidance, broad definitions, and inconsistent interpretation could raise costs or produce disclosure fatigue resembling Europe's cookie-banner experience.</cite> <cite index="24-4">Several member states have not yet stood up their market-surveillance authorities, so enforcement capacity will be patchy through 2026.</cite>