Analysis: US Government Raises Concerns about EU’s AI Act, Warns of Potential Negative Impact on Investment and Innovation

EU’s AI Legislation Agenda: Potential Implications and Controversies Surrounding the Proposed AI Act

The US government has raised concerns about the EU’s AI Act, warning that the legislation could hinder investment in AI technology and favor larger companies over smaller competitors. The analysis, conducted by the State Department, also suggests that the regulations could dampen productivity gains from AI and potentially lead to a migration of jobs and investment to other markets.

According to a report from Bloomberg, the analysis highlights some of the rules in the EU’s AI Act as “vague or undefined,” raising questions about the impact on AI research, development, and commercialization within the EU. The resource-intensive nature of training large language models (LLMs) could limit the competitiveness of European firms, with compliance costs potentially favoring only the largest tech companies.

While the US government has not yet published a comprehensive legislative plan for AI, it is taking a risk-focused approach to regulating AI models’ use. In contrast, the EU’s legislation is more focused on how AI models are developed.

The EU’s AI Act, which was approved by the European Parliament in July, includes requirements for transparency in generative AI systems and a ban on biometric surveillance in public settings and “social scoring” systems. However, some members of Parliament have raised concerns about the total ban on biometric use potentially hindering crime-solving and counter-terrorism efforts.

As the EU countries need to agree on the bill before it becomes law, the debate over the impact of the AI Act on innovation and competition in the AI sector is likely to continue. The State Department declined to comment on the leaked analysis, emphasizing the importance of digital solidarity between the US and the EU on key bilateral issues.