Lawmakers in the European Union (E.U.) recently approved legislation aimed at regulating artificial intelligence as the industry expands rapidly. The Artificial Intelligence Act (AI Act) was passed with an overwhelming majority of 523–46 votes, with 49 votes not cast. The legislation’s primary goal, as stated by the E.U. parliament, is to ensure safety, compliance with fundamental rights, and promote innovation. However, there are concerns that the law may impede innovation, especially considering the dynamic nature of the technology.
The law emphasizes the need for a defined risk-based approach to establish effective rules for AI systems. It categorizes AI systems into four groups, with high-risk systems facing bans, including those involved in manipulating human behavior or social scoring. Additionally, the use of biometric identification in public spaces for law enforcement purposes is restricted, with some exceptions.
High-risk AI systems, such as those used in critical infrastructure and public services, will undergo risk assessment and oversight. Transparency requirements will apply to limited-risk apps and general-purpose AI, while minimal-risk AI systems will remain unregulated.
In addition to managing risks, the law aims to create a governance structure at both European and national levels. The establishment of the European AI Office, serving as the central hub for AI expertise in the E.U., and an AI board as the primary advisory body on the technology are notable provisions.
Non-compliance with the law can result in significant penalties, ranging from €7.5 million to 7 percent of global revenue, depending on the infringement and company size. The regulation of AI will now be centralized across E.U. member states to establish a harmonized standard for such oversight.
While the E.U. is not alone in enacting AI regulations, with China and the U.S. also taking steps in this direction, critics express concerns that the stringent rules may stifle innovation and competition, potentially limiting consumer choice. The law’s focus on risk management has drawn criticism for potentially hindering technological advancements without fully considering the benefits.
Despite the criticisms, certain aspects of the law, such as restrictions on biometric identification, are seen as positive steps towards safeguarding civil liberties. However, the law’s numerous exceptions for national security concerns have raised some eyebrows.
As the U.S. considers its own approach to AI regulation, inspired by the E.U.’s AI Act, the balance between fostering innovation and protecting civil liberties remains a key consideration. It is essential for lawmakers to strike a balance that encourages innovation while ensuring the protection of individual rights.
