Estonia is set to incorporate European Union transparency standards into national law, requiring clear disclosure of AI-generated content and measures to prevent misleading synthetic media, aiming to bolster trust and accountability across sectors.
Estonia is preparing to bring new European Union transparency rules on artificial intelligence and deepfakes into domestic practice, in a move officials say is designed to make it harder for people to be misled by synthetic content. The rules require clearer disclosure when people are dealing with an AI system or viewing material generated by AI, and they are intended to apply across sectors rather than only in narrow technical fields, according to Estonia’s Ministry of Justice and Digital Affairs and the European Commission.
Aivar Pau, a strategic communications adviser at the ministry, said the concern is that AI can present itself in a way that looks human, or be embedded in a service without being obvious to the user. In those cases, people may wrongly assume they are speaking to a real person. The ministry said disclosure is not needed when the AI use is self-evident, such as in a clearly labelled chatbot or AI character, or where the design of the service makes that role plain.
The rules also extend to biometric and sensitive categorisation tools. Pau said users must be told when systems are capable of processing biometric data, inferring emotions or intentions, or classifying people by gender, age, appearance, ethnic origin or personal preferences. The European Commission has also published three visual labels for AI use: a basic icon, a mark for fully AI-generated content and a mark for content partly created with AI. Those labels are meant to stay attached throughout the creative process.
Deepfakes receive separate treatment. Content created or altered by AI and likely to be mistaken for authentic material must be identified, although there are exceptions for law-enforcement use and for edited texts that have been reviewed by a responsible human editor. The ministry said any required information must also be accessible to people with disabilities. ERR reported that violations may carry fines of up to €15,000, while the EU AI Act itself sets a wider enforcement ceiling of up to €15 million or 3% of a company’s worldwide annual turnover. Estonia will still need to transpose the rules into national law before local authorities can be formally tasked with supervision and penalties.
Disclaimer: This content is intended for informational purposes only. Readers are advised to exercise their own judgement, conduct due diligence, or consult a qualified expert before acting on any information provided.
