Prague – The new provisions of the Artificial Intelligence (AI) Act, which took effect on Saturday, will primarily impact major AI model providers like ChatGPT, Gemini, and Claude. Experts believe the regulations are overdue, with ongoing ambiguity surrounding their interpretation until the last minute, as highlighted by statements to ČTK. The AI Act was established in August of the previous year, marking the first comprehensive AI regulation globally. It was accompanied by a voluntary code aimed at limiting copyright violations in content generation and introducing risk assessment mechanisms prior to and following a model’s market introduction.
In early July, lawyer Štěpánka Havlíková from Dentons’ Prague office noted significant uncertainty in interpreting the AI Act rules. “Following extensive discussions, the Code of Good Practice for General AI Models was released, as directly anticipated by the AI Act. This code clarifies how to meet the obligations specified in the AI Act and aids providers in demonstrating compliance,” she remarked.
“Although the code is not legally binding and participation is voluntary, it will greatly assist providers in proving adherence to transparency, copyright, and safety standards set by the AI Act,” Havlíková stated. Providers opting out of signing the code must show they have implemented suitable, effective, and proportional measures to comply with the obligations of the act.
According to Daniil Shakhovsky, co-owner of Lexicon Labs, voluntary codes alone are inadequate. “Procrastination leaves us reliant on the models of larger players. Those lacking tools to verify content authenticity are at a disadvantage. We need mandatory minimum standards and infrastructure that operates independently of platform oversight. Action is rewarded, not inaction,” he asserted.
The code’s publication is an important milestone, as per Lukáš Benzl, director of the Czech Association of AI, though it arrives fairly late. Providers of general AI models faced uncertainty about European requirements until the last moment. “Such uncertainty complicates planning, stifles innovation, and erodes confidence in regulatory stability. It is essential to rethink the pace and complexity of regulations, especially considering the AI action plans of the USA and China. If the EU aspires to be a global leader, it must excel not only in ambition but also in the quality and predictability of its processes,” Benzl concluded. (August 3)
Leave a Reply