On April 29, local time in Brussels, after a 12-hour marathon negotiation between EU member states and the European Parliament, they still failed to reach an agreement on a package of "weakened" artificial intelligence rules and could only postpone the discussion until next month. The relevant modifications revolve around the Artificial Intelligence Act (AI Act), which will officially take effect in August 2024. The key provisions of the Act were originally planned to be implemented in phases starting this year.
This adjustment is part of the "Digital Omnibus" (Digital Omnibus) promoted by the European Commission, which is intended to simplify a number of regulatory rules in the digital field and help European companies catch up with American and Asian rivals in the artificial intelligence competition.
Europe's current AI regulatory framework is generally regarded as the most stringent set of rules in the world. The legislative background is rising concerns from all walks of life about the impact that artificial intelligence technology may have on children, workers, businesses and network security. Cyprus, which currently holds the rotating EU presidency, stated that it “cannot reach an agreement with the European Parliament.” Kim van Sparrentak, a Dutch member of the European Parliament, sharply criticized the failure of the negotiations, saying, "Tech giants are probably popping the champagne to celebrate, while European companies that value safety and have prepared in accordance with regulations are now facing regulatory chaos."
The next round of negotiations is expected to take place in two weeks, according to sources with direct knowledge of the negotiations. They said that the current round of negotiations, which started at 11:00 GMT on Tuesday, was blocked mainly because some member states and some parliamentarians insisted that industries already subject to sectoral regulations (such as areas covered by product safety rules) should be excluded from the scope of AI legislation.
Current AI regulatory requirements impose stricter compliance obligations on application scenarios considered "high risk", including biometric identification, utility supply, medical health, credit assessment, and law enforcement activities. At the same time, the digital comprehensive revision plan also involves many important regulations such as the General Data Protection Regulation (GDPR), the e-Privacy Directive and the Data Act. The proposed changes to the above-mentioned regulations and the AI Bill have triggered strong criticism from privacy defenders and civil rights groups, who believe that the EU's move is a "concession" to technology giants.
Related articles:
New EU regulatory trends: The scope of the Digital Market Act will be extended to cloud services and AI