EU has issued new rules that generative AI tools must disclose copyright information.

CTOnews.com, April 28 (Xinhua) under an early EU agreement, companies deploying generative artificial intelligence tools such as ChatGPT will have to disclose any copyrighted material used to develop their systems, which could pave the way for the world's first comprehensive law governing the technology.

The European Commission began drafting the artificial Intelligence Act two years ago to regulate this emerging technology, which became a craze after OpenAI released ChatGPT, Reuters reported. Members of the European Parliament have agreed to take the draft to the next stage, a tripartite dialogue, when EU lawmakers and member states will finalize the final details of the bill.

Under the proposal, the AI tool would be classified according to its perceived level of risk: from lowest to limited, high and unacceptable. Areas of concern include biometric monitoring, the dissemination of false information or discriminatory language. CTOnews.com notes that while high-risk tools are not prohibited, people who use these tools need to be highly transparent in their operations.

Companies that use generative AI tools must also disclose any copyrighted material used to develop their systems. The provision is a later supplement drawn up in the past two weeks, according to a person familiar with the matter. Some committee members initially proposed a total ban on the use of copyrighted materials to train generative AI models, but the proposal was abandoned in favour of transparency requirements, the person said.

"with conservatives wanting more surveillance and leftist fantasies of over-regulation, parliament has found a solid compromise that can moderately regulate artificial intelligence, protect citizens' rights, and promote innovation and economic development," said Svenja Hahn, a deputy member of the European Parliament.