The European Commission has published guidelines to assist providers and deployers of artificial intelligence (AI) systems in complying with the transparency obligations of the European Union’s incoming AI Act, which will come into force on August 2, 2026.
The AI Act’s transparency obligations are aimed at helping people recognize when they are interacting with AI or when content has been generated or altered by AI, “reducing the risk of deception and manipulation.”
The new guidelines clarify which providers and deployers must comply with the transparency obligations for interactive AI systems and the marking and labeling of AI-generated content.
“With today’s guidelines, the Commission supports the smooth and effective application of the AI Act to make AI systems interacting with people such as chatbots and AI agents and AI content more transparent and trustworthy,” Henna Virkkunen, Executive Vice-President of the European Commission for Tech Sovereignty, Security and Democracy said. “These guidelines support providers and deployers in meeting their obligations under the AI Act, while helping citizens know when they are interacting with AI.”
In a July 20 statement, the Commission explained that the drafting of the new guidelines was informed by input from a variety of stakeholders, including EU member states in the European AI Board, a key advisory body created by the AI Act.
Transparency rules
Under the AI Act, AI providers will have to design AI systems to inform users when they are directly interacting with AI, as well as add machine-readable marks to enable the detection of AI-generated or manipulated content.
Deployers will also have to inform users when they are exposed to deepfakes, AI-generated content on matters of public interest without human review or editorial control, and emotion recognition or biometric categorization systems.
“The guidelines explain certain concepts, give exemptions and examples,” said the Commission. “These include what constitutes directly interactive AI systems, such as chatbots, synthetic content, including partially or fully AI-generated text, deepfakes, and AI-generated text on matters of public interest.”
Finally, the guidelines explain how compliance with the AI Act’s transparency obligations may be demonstrated, including by adhering to a “Code of Practice,” which provides legal certainty and “a simple and practical way to demonstrate compliance with the AI Act.”
AI Act arrives
The AI Act was passed by the EU Parliament in March 2024, aiming to introduce a regulatory framework that categorizes AI applications by perceived risk.
Under the Act, lower-risk applications, such as spam filters and content recommendation systems, are subject to minimal regulations, including a requirement to disclose their use of AI to ensure transparency. Meanwhile, high-risk AI systems, particularly those deployed in sensitive sectors such as healthcare, education, and public services, face more stringent regulatory requirements, including detailed documentation of their processes and mandatory human oversight of their operations.
The Act also imposes outright bans on certain AI applications, such as social scoring systems, predictive policing, and emotion recognition systems in schools and workplaces, which are prohibited due to their potential to infringe on individual freedoms and rights. In addition, the use of AI for biometric identification by police in public spaces is restricted, except in cases involving serious crimes such as terrorism or kidnapping.
The legislation entered into force on August 1, 2024, and will be fully applicable two years after this date, meaning August 2, 2026. Prohibited AI practices and AI literacy obligations entered into application earlier, on February 2, 2025.
Since its passage into law, the AI Act has been revisited by EU lawmakers to provide additional clarity.
In May, the European Council and European Parliament reached a provisional agreement on a proposal to “streamline” certain rules, including delaying their implementation, reducing requirements for certain businesses, and increasing focus on reducing harm to children.
The proposal to change certain aspects of the Act formed part of the EU’s so-called “Omnibus VII” legislative package, put forward by the Commission to simplify the EU’s digital legislative framework.
In terms of next steps, after August 2, most of the rules laid down in the AI Act will start to apply, including the enforcement powers of the Commission and of national market surveillance authorities. The Commission also noted that AI systems placed on the market before the August date will have until December 2, 2026, to get compliant with the marking and detecting obligations.
To help organizations better understand their obligations, the Commission put in place a range of tools, including guidelines, such as those newly published on transparency, the Code of Practice, and the AI Act Service Desk, an accessible, up-to-date information hub offering clear guidance on the AI Act.
In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.
Watch: AI is becoming a game-changer for iGaming
Be the first to comment