STOCKHOLM, April 27 (Reuters) – Companies that deploy generative AI tools such as ChatGPT will have to disclose any copyrighted material used to develop their systems, according to an early European Union agreement that could pave the way for the world’s first sweeping laws. technology control.
The European Commission began drafting AI law nearly two years ago to regulate the emerging technology, which saw a surge in investment and popularity following the release of OpenAI’s AI-powered ChatGPT chatbot.
MEPs agreed to push the draft into the next stage, the triennium, during which lawmakers and member states will discuss the final details of the bill.
Under the proposals, AI tools will be categorized according to their perceived level of risk: from minimal to limited, high and unacceptable. Areas of concern could include biometric surveillance, spreading misinformation or discriminatory language.
While high-risk tools will not be banned, those who use them will need to be very transparent with their operations.
Companies that deploy generative AI tools, such as ChatGPT or image generator Midjourney, are required to disclose any copyrighted material used to develop their systems.
This provision was a late addition that was drawn up over the past two weeks, according to a source familiar with the discussions. Some panelists initially proposed banning the use of copyrighted material to train generative AI models altogether, the source said, but that was dropped in favor of a transparency requirement.
Svenja Hahn, MEP, said: “Against conservative desires for more surveillance and left-wing fantasies of over-regulation, Parliament has reached a powerful compromise that would regulate AI proportionally, protect citizens’ rights, as well as foster innovation and boost the economy. “. .
Microsoft-powered OpenAI (MSFT.O) sparked fear and anxiety around the world when it unveiled ChatGPT late last year. The chatbot has become the fastest growing consumer app in history, reaching 100 million monthly active users in a matter of weeks.
The ensuing race among tech companies to bring generative AI products to market has some onlookers worried, with Twitter owner Elon Musk backing a proposal to halt development of such systems for six months.
Shortly after signing the letter, the Financial Times reported that Musk was planning to launch his own company to compete with OpenAI.
(Reporting by Suvantha Mukherjee in Stockholm and Fu Yun Che in Brussels
Our standards: Thomson Reuters Trust Principles.