BRUSSELS, July 31 (Xinhua) -- The European Commission said on Friday that it will begin enforcing a broader range of provisions of the European Union(EU)'s Artificial Intelligence (AI) Act on August 2, when new transparency requirements for AI systems also take effect.
Under the rules, chatbots and other interactive AI systems must inform users that they are interacting with AI rather than a human. AI-generated or altered images, videos and audio, including deepfakes, must be clearly labelled and carry machine-readable markings.
The requirements are aimed at reducing deception and manipulation while helping users identify AI-generated content, the Commission said.
The Commission's AI Office will enforce rules covering providers of general-purpose AI models, including advanced models that may pose systemic risks. Providers will be required to document relevant information, adopt copyright policies and publish sufficiently detailed summaries of the content used to train their models.
Providers of the most advanced models must also take measures to address risks linked to cybersecurity, harmful manipulation, loss of human control and threats to fundamental rights.
Enforcement will also begin for prohibited AI practices, including systems that manipulate people, exploit vulnerabilities or conduct certain forms of social scoring.
The AI Office, national competent authorities and the European Data Protection Supervisor will share responsibility for enforcement.
The Commission said more than 180 organizations have signed a voluntary code of practice designed to help companies comply with the transparency requirements.
Under the AI Omnibus, rules for certain high-risk AI systems have been postponed until December 2027, while requirements for high-risk systems integrated into regulated products will apply from August 2028.
The EU AI Act, which entered into force on Aug. 1, 2024, is being implemented in phases. ■