The European Parliament has taken a significant step towards reining in the power of artificial intelligence and ChatGPT, as it voted Thursday, May 11, to approve pioneering legislation that would establish stringent regulations for these technologies. The new rules primarily target the use of facial recognition, biometric surveillance and other AI applications such as ChatGPT.
“This vote is a milestone in regulating AI, and a clear signal from the Parliament that fundamental rights should be a cornerstone of that,” Kim van Sparrentak, member of the European Parliament, told Reuters. “AI should serve people, society, and the environment, not the other way around.”
Thursday’s vote brings Europe closer to implementing a comprehensive ban on the utilization of facial recognition in public spaces. If the legislation is fully enacted, it would mean that neither pictures nor video footage captured in public venues, whether in real-time or afterwards, could be employed to identify individuals.
Under this proposed bill, AI tools would be classified based on their perceived level of risk, ranging from low to unacceptable. Subsequently, companies and governments would receive specific guidelines to follow, depending on the risk level associated with the AI technologies they utilize.
“It is a delicate deal. But it is a package that I think gives something to everyone that participated in these negotiations,” said Dragos Tudorache, one of the parliamentarians charged with drafting the laws. “Our societies expect us to do something determined about artificial intelligence, and the impact it has on their lives. It’s enough to turn on the TV … in the last two or three months, and every day you see how important this is becoming for citizens.”
While AI technologies and capabilities are rapidly advancing, Europe is not traditionally considered a stronghold in AI development, as the United States and China currently dominate this field. However, the United States currently lacks federal legislation regulating AI, though the Biden administration has introduced an “AI Bill of Rights” to establish certain parameters, but they remain non-binding and have no legal enforceability.
