The current AI boom, from Bing and Midjourney, relies on free access to training data, much of which is culled from the web and often protected by copyright. The use of this data has led to criticism and lawsuits, particularly in the art world, where rights holders have argued that their works are being exploited without their permission.
Some of the world’s biggest players in artificial intelligence, such as OpenAI, have avoided scrutiny by denying details of the data used to create their software. But proposed legislation in the EU to regulate artificial intelligence (the Long-Term and Far-reaching Artificial Intelligence Act) could force companies to disclose this information, according to reports from Reuters And uractive.
The amendment was reportedly a late addition to the AI bill
Reuters He says late amendments to the AI law, which were approved in draft form by lawmakers earlier this week, would require “companies that publish Al-creational tools, such as ChatGPT … to disclose any copyrighted material used to develop their systems.” earlier this month, uractive It reported on the same ruling, saying that companies should “make available to the public a summary disclosing the use of training data protected by copyright law.” Reuters Citing “sources familiar with the discussions,” he said the amendment was a “late addition that has been put in place over the past two weeks.”
Details of this requirement are not known, and the law may change during upcoming closed-door negotiations, known as triads, needed to finalize the law. But if AI companies are forced to disclose the sources of their training data, it could open the door to many lawsuits that could affect some of the biggest names in technology.
Already, companies like Getty Images are suing image-generating AI for scraping their data without permission, while there are a few class-action lawsuits targeting the AI that creates the images and the encryption. Yet the biggest name in artificial intelligence today—OpenAI, maker of ChatGPT, GPT-4, and DALL-E and the force behind Microsoft’s AI push—is very secretive about its data sources. Reported legislation could change this, providing evidence for lawsuits and leverage for discussions between organizations such as media companies, whose data is used and referenced by many chatbots.
Although the potential impact of the law will depend on its details, the rest of EU AI law is certain to have similarly broad effects on the rapidly changing AI landscape.
The law would classify AI systems based on their perceived risk and would require companies responsible for building the most impactful tools to disclose critical data about safety, interpretation, performance, and so on. As with previous tech regulations pushed by the European Union, the AI law will undoubtedly have a global impact on how tech companies do business. EU lawmakers will continue to debate the details of the law throughout the year, though compliance for companies likely won’t come into force until 2025 or later.