After three days of intense negotiations, EU member states and European Parliament lawmakers reached an agreement on Friday regarding a framework to regulate artificial intelligence. This groundbreaking text is the first of its kind globally and is expected to foster innovation in Europe while mitigating potential abuses of advanced technologies. The European commissioner who initiated the project in 2021 has described the agreement as ‘historic’.
EU recognizes the need for AI regulating legislation
The lengthy discussions lasting up to 37 hours in the last round highlight the division among EU member states on how to handle AI regulation. The process, initiated in 2021, faced disruptions due to the launch of ChatGPT at the end of 2022, showcasing the progress of generative AI and associated risks like false information and deep fakes.
At the request of EU lawmakers, a specific framework for generative AI was created, causing additional delays. This situation reflects the common struggle of laws to keep pace with technological advancements, with differences among EU member states complicating matters.
While the agreement marks a significant step forward, it still requires formal approval from member states and the Parliament. Enforcement is anticipated by 2025, allowing room for further developments in the AI domain. The two-tier mechanism of the law imposes different obligations on AI system providers based on the power and perceived riskiness of their systems.
For AI models like ChatGPT, obligations include risk management, monitoring human control, and technical documentation. Less powerful algorithms are subject to transparency requirements. The law also addresses the use of AI by law enforcement, prohibiting real-time facial recognition and automatic biometric identification in public places to prevent a social credit system and the creation of an environment of constant surveillance. However, certain exemptions, such as the prevention of terrorist threats and identification of suspected criminals, raise concerns.
The political agreement reached on Friday is a significant development, but the formal approval process and enforcement are yet to be completed. The law aims to strike a balance between promoting innovation and regulating the potentially risky aspects of AI.
Comments 3