An early EU agreement says that companies that use generative AI tools like ChatGPT will have to reveal any copyrighted materials they used to build their systems. This could lead to the world’s first comprehensive rules governing technology.
Nearly two years ago, the European Commission started writing the AI Act to control new artificial intelligence technology. After OpenAI‘s AI-powered chatbot, ChatGPT came out, there was a big increase in investment and interest in the field.
Members of the European Parliament agreed to move the bill to the next stage, called the “trilogue,” where EU lawmakers and member states will work out the bill’s final details.
Under the plans, AI tools will be put into four groups based on how dangerous they are thought to be: minimal, limited, high, and unacceptable. Concerns could include biometric monitoring, spreading false information, or using language that makes people feel bad about themselves.
Even though high-risk tools won’t be banned, those who use them will have to be very open about how they work.
Companies that use generative AI tools like ChatGPT or the picture generator Midjourney will also have to share any copy-protected information they used to build their systems.
A person familiar with the talks said that this rule was added at the last minute, in the last two weeks. The source said that at first, some committee members wanted to ban copyrighted material from being used to train generative AI models, but this idea was dropped in favor of a transparency rule.
“Conservatives wanted more surveillance, and the left wanted too much regulation,” said Svenja Hahn, a member of the European Parliament. “But parliament found a good middle ground that would regulate AI in a fair way, protect citizens’ rights, encourage innovation, and help the economy,” she said.
Fred Havemeyer, a Macquarie expert, said that the EU’s plan was “tactful” and not like the “ban first, ask questions later” plan that some people had suggested.
“The EU has been at the forefront of figuring out how to regulate AI,” he told Reuters.
Race to market
Supported by Microsoft (MSFT.O) When OpenAI showed off ChatGPT at the end of last year, people all over the world were amazed and worried. The robot grew faster than any other consumer app in history, going from 0 to 100 million monthly active users in just a few weeks.
Some people were worried about the race between tech companies to bring creative AI products to market. Elon Musk, who owns Twitter, supported a plan to stop developing such systems for six months.
Shortly after Musk signed the letter, the Financial Times said that he was going to start a company to compete with OpenAI.