E.U. lawmakers call for a different set of regulations for A.I. tools like ChatGPT

April 18, 2023: A firm of a dozen lawmakers for the European Union known for a new set of rules to regulate an enormous swath of artificial intelligence tools beyond those recognized as explicitly high risk under the region’s proposed A.I. Act.

On Monday, in an open letter, the members of parliament called for international cooperation in guardrails on the development of A.I. tools, which industry experts have cautioned could become dangerously powerful without proper precautions.

The letter comes after a small group of prominent A.I. experts called for Europe to make its A.I. rules more expansive, arguing that excluding general purpose A.I., or GPAI, would miss the mark. General purpose A.I. includes broad tools like ChatGPT that might not be designed with an increased-risk use in mind but could be utilized in settings that elevate their risks.

In the open, lawmakers acknowledged that while the A.I. Act is directed at high-risk A.I. use cases, “we also need a complementary set of preliminary rules for developing and deploying powerful General Purpose A.I. systems that can be easy to a multitude of purposes.”

The lawmakers reference one more recent letter from the Future of Life Institute, signed off by billionaire Elon Musk, Co-founder of Apple Steve Wozniak and the 2020 presidential candidate Andrew Yang, from many others. In that letter, technology leaders for a minimum six-month pause on training A.I. systems more potent than OpenAI’s most recent large language model, GPT-4.

“We share some of the concerns shown in this letter, even while we disagree with a few of its more alarmist comments,” the members of parliament wrote.

“We are nevertheless with the letter’s core message: with the rapid evolution of powerful A.I., we see the need for huge political attention.”

They pledged to give a set of rules within the A.I. Act framework to steer A.I. development in a “human-centric, safe, and trustworthy” way. They are also known as Ursula von der Leyen and U.S. President Joe Biden convening a Global Summit on A.I. where they permit preliminary which governs principles for the “development, control, and deployment” of A.I.

The lawmakers said democratic and non-democratic countries should be called “to exercise restraint and responsibility in their pursuit of compelling artificial intelligence.” And they advocated for A.I. labs and companies to maintain a sense of commitment and to surge transparency and discussion with regulators.