The European Union’s draft law to regulate artificial intelligence (AI) must be updated to include overarching controls on chatbots like ChatGPT, a group of European Parliament members wrote in an open letter shared Monday.

The European Union has been out front globally in its effort to regulate AI. It was expected the AI Act, which was proposed by the European Commission in 2021 and has been under debate since, would be voted on in the coming months.

But the act was written before the release of ChatGPT by OpenAI in November and other generative AI tools, which can masquerade as human intelligence and converse with adults and children.

The “widespread public access” to sophisticated, AI-generated content “has prompted us to pause and reflect on our work,” wrote the group of 12 lawmakers including Dragoș Tudorache of Romania and Brando Benifei of Italy, lead authors of the AI Act. Tudorache posted the text of the open letter on his Twitter account.

The AI Act as currently written covers only specific, high-risk cases of AI, the lawmakers wrote. The law needs to include “preliminary rules for the development and deployment of powerful, general purpose AI systems that can be easily adapted to a multitude of purposes,” they said, including controls on developers and users of generative AI.

The lawmakers said they took note of a recent letter written by the Future of Life Institute and co-signed by thousands of experts and business leaders calling for a six-month moratorium on the development of powerful AI.

“We are … in agreement with the letter’s core message: with the rapid evolution of powerful AI, we see the need for significant political attention,” the lawmakers said.

The letter called for European Commission President Ursula von der Leyen and U.S. President Joe Biden to convene a global summit on AI, with the goal of agreeing to a “preliminary set of governing principles for the development, control, and deployment of very powerful artificial intelligence.”

ChatGPT’s capabilities have raised alarm bells worldwide, with the platform swiftly banned in China, Iran, North Korea, and Russia.

The Italian data protection authority (DPA) recently shut down ChatGPT in the country, alleging the platform has no controls to stop it interacting with children. The European Data Protection Board said it would form a task force to coordinate the potential enforcement actions of other DPAs, after data privacy regulators in France, Germany, Ireland, and Spain each indicated they might consider similar action.

Canada is also investigating possible improper handling of personal data by ChatGPT, while the U.S. Commerce Department is seeking feedback on mechanisms to help better create trust in AI.

Adrianne Appel writes regulatory news, policy, and trends for Compliance Week. She previously reported about policy developments for Bloomberg Law and Bloomberg Government.

Topics