EU begins enforcing AI Act as transparency rules for chatbots and deepfakes take effect

The European Commission will start enforcing key provisions of the Artificial Intelligence (AI) Act on 2 August 2026, marking a decisive shift from regulation to practical oversight of one of the world’s most ambitious legal frameworks for artificial intelligence. On the same date, new transparency requirements enter into force, obliging certain AI systems to inform users when they are interacting with artificial intelligence and when content has been generated or altered by it.

Under the new rules, chatbots and other interactive AI systems must clearly disclose that users are dealing with AI rather than a human. Deepfakes – images, videos or audio edited or generated using AI – must be labelled. AI-generated or altered content will also have to carry machine-readable marks to facilitate detection. The measures are designed to reduce deception and manipulation and to help citizens make informed choices, while giving businesses clearer obligations and a practical route to demonstrate compliance.

The Commission published a first list of more than 180 organisations that have signed the Code of Practice on transparency of AI-generated content, which operationalises these rules.

Enforcement of general-purpose AI models

The AI Office can now enforce the Act’s rules for providers of general-purpose AI (GPAI) models – systems capable of performing a wide range of tasks and powering tools and services, including AI agents. The most advanced models that may pose systemic risks face additional obligations to address large-scale harms, including risks linked to chemical, biological, radiological and nuclear incidents, loss of control, cyber offence, harmful manipulation, threats to fundamental rights, European cybersecurity and AI systems acting outside human control.

All GPAI providers must document specified information and make it available to competent authorities or downstream providers. They are also required to put in place a copyright policy and publish a sufficiently detailed summary of the content used to train their models.

Enforcement also begins for transparency obligations and prohibited AI practices. The latter ban particularly harmful systems, including those that manipulate people, exploit vulnerabilities in harmful ways, or unfairly score individuals in ways that threaten their rights.

Responsibility is shared. The AI Office enforces the rules for AI systems offered by the same provider as the underlying general-purpose model, as well as systems integrated into very large online platforms or search engines designated under the Digital Services Act. National competent authorities handle other AI systems, while the European Data Protection Supervisor oversees systems used by EU institutions, bodies and agencies. Effective enforcement will depend on member states properly designating and resourcing their national authorities.

Scientific support and monitoring tools

Enforcement will be backed by a Scientific Panel of 60 independent AI experts, which has already held its first meeting. The AI Office has appointed Professor Alessandro Abate of the University of Oxford’s Department of Computer Science as Lead Scientific Adviser. He will support work on general-purpose AI models, including innovation, adoption, testing and evaluation.

To facilitate oversight, the AI Office has launched several tools. Natural and legal persons can use a Complaint Tool to report alleged infringements by providers supervised by the Office. A Whistleblower Tool allows people working with AI providers to report possible violations securely. A dedicated channel is available for downstream providers building systems on general-purpose models who wish to report alleged infringements by the model providers. Information received through these channels will be treated confidentially.

Significance and expected impact, including on media

The start of enforcement represents a landmark moment for the EU’s risk-based approach to AI. Europe positioned itself early with a comprehensive framework intended to give innovators legal certainty while protecting the public interest. As Executive Vice-President for Tech Sovereignty, Security and Democracy Henna Virkkunen stated: “AI is a transformative technology that can bring extraordinary benefits to our people and businesses. But we are also seeing that harms can occur if AI is not properly designed and used and the most advanced models create risks on an entirely new scale. […] As enforcement begins, we are taking an important step towards AI that people and businesses can understand and trust, and whose benefits are shared widely across our society.”

For the media sector the transparency rules carry particular weight. The mandatory labelling of deepfakes and the requirement for machine-readable marks on AI-generated or altered content are expected to strengthen the ability of journalists, platforms and fact-checkers to identify synthetic media, a growing challenge for newsrooms confronting sophisticated disinformation campaigns. Clear disclosure when users interact with AI systems may also affect how news organisations deploy chatbots, automated content tools and recommendation systems, requiring greater transparency toward audiences.

The obligations on GPAI providers – including copyright policies and summaries of training data – could influence the tools many media outlets rely on for research, writing assistance and content generation, potentially shaping licensing practices and liability debates around AI-assisted journalism.

Next steps

The AI Omnibus has postponed the application of rules on high-risk AI systems to 2 December 2027 and those for high-risk systems integrated into regulated products to 2 August 2028. New prohibitions on AI systems that generate non-consensual sexually explicit content and child sexual abuse material will apply from 2 December 2026.

With enforcement now underway, the coming months will test the capacity of the AI Office, national authorities and the Scientific Panel to turn the Act’s ambitions into consistent, effective oversight across the single market.

Caption: A cheeky AI-generated illustrative image.