Brussels: New transparency rules for general-purpose artificial intelligence (AI) models came into effect today across the European Union as part of the bloc's landmark AI Act, adopted last year. The rules aim to promote the responsible development of AI, provide clearer information on how models are trained, strengthen copyright protection, and require developers to outline the steps taken to safeguard intellectual property.

According to Qatar News Agency, under the new rules, developers must disclose how their models work, the data used for training, and document safety measures, particularly for highly advanced models that are considered to pose potential risks to the public. General-purpose AI refers to systems that can be applied to a wide range of uses, such as generating text, analyzing language, or writing code.

General models are defined as those with training capacity exceeding 10^23 floating-point operations (FLOPs) and capable of generating language. Highly advanced models, exceeding 10^25 FLOPs and posing systemic risks, face additional obligations, including notifying the European Commission and ensuring safety and security measures are in place.