From August 2, new rules come into effect in the EU imposing greater transparency on AI systems

Starting from August 2, a new phase of the EU AI Act will come into force, introducing additional transparency obligations for providers and users of systems based on artificial intelligence. The new provisions will require that users are clearly informed when interacting with a chatbot, an AI agent, or any other automated system designed to communicate with people.

In addition to the obligation to identify interactions with artificial intelligence, the rules stipulate that content generated or modified through AI must be accompanied by machine-readable information, making its origin recognizable. The stated aim of the European Commission is to strengthen user trust and promote a more transparent use of new technologies.

In presenting the guidelines ahead of the new rules coming into effect, Henna Virkkunen, Executive Vice President of the European Commission responsible for technological sovereignty, security, and democracy, explained that the guidance has been developed to help businesses and organizations properly apply the obligations set forth in the AI Act, while providing citizens with the tools to understand when they are interacting with an artificial intelligence system.

New EU Provisions on AI Use

The new provisions cover numerous categories of applications. These include chatbots, virtual assistants, audio, images, videos, and text generated artificially, as well as deepfakes. The rules also extend to texts produced or modified by artificial intelligence intended to inform the public on matters of general interest, unless such content has undergone actual human review or editorial control.

Also affected are systems dedicated to emotion recognition and those used for the biometric categorization of people. However, tools performing simple editing tasks, such as automatic correction of grammatical or spelling errors, are excluded as long as they do not substantially alter the original content.

The AI Act represents the first comprehensive regulatory framework dedicated to artificial intelligence adopted by the European Union. Officially coming into effect in 2024, the regulation has been introduced progressively. Some provisions are already operational, while others will be applied in the coming years.