The European Parliament is working through key legislation to regulate the use of artificial intelligence to regulate its use based on its ability to do harm to society and business, including measures that would force chatbots to declare they’re not human.

Members of the European Parliament are meeting currently in Brussels over a set of proposals to form part of what will become Europe’s Artificial Intelligence Act, which is a set of regulations regarding the use of AI, the Financial Times reported today.

Among the proposals that will likely enter the regulation are the requirement for chatbots to inform users that they’re not real people. That way they cannot accidentally trick users into thinking they are talking to another person. The MEPs also want developers of AI products, such as OpenAI LP’s ChatGPT to declare if the product uses copyrighted information to train their AI models.

Training data can be a big issue, because it’s often collected by AI providers without the permission of content creators, explained ImmunoWeb SA founder Ilia Kolochenko. “While modern intellectual property law provides from little to no protection to copyrighted content,” he said, “most large-scale data-scrapping practices likely violate terms of service of digital resources, such as online libraries and websites, and may eventually lead to an avalanche of litigation for breach of contract and interrelated claims.”

If an AI happens to scrape the internet and use copyrighted materials, a measure would be added that would permit copyright holders to get paid for their work as it’s used in the AI. That follows the recent lawsuit against AI art generators Stability AI and Midjourney that alleged the AIs have used copyrighted works “without the consent of the original artists.” Getty Images also filed a lawsuit against Stability AI, alleging it used more than 12 million copyrighted images in its training set.

Brando Benifei, an Italian politician and member of parliament, said that he wants even further transparency on the subject. “We want a list of the public disclosure of the material that is being used to train it because [authors] can go through other legislations and avenues to try to get paid for what is being used without their consent,” he said.

A copy of the draft legislation, viewed by Euractiv, could also include a potential ban on AI use for predictive policing, which is where algorithms are used to guess what crimes might happen or identify suspects before anything happens, and a ban on emotion recognition, which is detecting feelings and state of mind. That ban would also extend to workplace and educational institutions, with an exception for medical and research where consent is given.

The talks over this legislation follow the temporary ban of OpenAI’s ChatGPT services by regulators in Italy over privacy concerns and began a probe into its practices. On Wednesday, Italian regulators outlined rules for the reinstatement of OpenAI’s services. The European Data Protection Board, an independent privacy watchdog group in the European Union, formed its own task force to investigate ChatGPT.

“Of note, the regulatory trend is not a prerogative of European regulators, for example in the United States the FTC is poised to actively shape the future of AI,” added Kolochenko. “The Cyberspace Administration of China is also energetically working on new rules and restrictions for AI companies.”

Much of the world is gearing up regulatory attention on the use of AI, including the U.S. and China, as both countries are seeking public comment to craft future rules regarding the technology.

It’s too soon to say what may go into the text of the legislation, but the end of discussions is expected to finalize next week, potentially before a key committee vote on April 26.

Image: geralt/Pixabay