US Raises Concerns Over EU AI Regulations, Fearing Impact on Smaller Firms and Competitiveness
The United States has raised concerns over the European Union’s proposed regulations on artificial intelligence (AI), warning that they could favor larger companies with the financial resources to comply with the law. Previously undisclosed documents obtained by Bloomberg News reveal the US analysis of the European Parliament’s version of the AI Act, which includes regulations on generative AI. According to the documents, some of the rules in the parliament law are based on vague or undefined terms.
The US analysis is a detailed position on the EU legislation and could set the tone for other countries writing their own rules for AI. One of the main concerns expressed by the US is that the European Parliament focuses on how AI models are developed, while the US prefers an approach that focuses on the risks associated with how these models are used.
The US warns that the proposed EU regulations could dampen productivity and potentially lead to job losses and investment migration to other markets. The new rules could also hinder investment in AI research and development (R&D) and commercialization in the EU, limiting the competitiveness of European firms. This is because training large language models, which are essential for generative AI products, requires significant resources.
The US State Department shared its feedback, including line-by-line edits of certain provisions in the AI Act, with European counterparts in recent weeks. The comments were intended to foster cooperation and alignment of values between the US and the EU. Some EU member countries have raised similar concerns about the European Parliament’s version of the AI Act.
The European Parliament’s AI Act, voted on in June, calls for more transparency regarding the source material used to train large language models. This vote paves the way for negotiations among parliament, the European Commission, and member states, with the goal of finalizing the rules by the end of the year.
The US analysis aligns with the State Department’s preference for a more hands-off approach to AI regulation in order to promote innovation. However, US officials have shown mixed views on regulation, with some becoming more open to mandatory rules as concerns about the potential harms of AI technology grow.
Aaron Cooper, head of global policy at BSA The Software Alliance, emphasizes the importance of countries reaching common ground on the basics of AI regulation. Cooper suggests that the Biden administration should continue engaging in open and candid discussions with European counterparts to align objectives for AI policy.
The concerns raised by the US highlight the potential impact of AI regulations on smaller firms and the competitiveness of European companies. As the EU moves forward with its AI regulations, balancing the need for innovation with adequate safeguards will be crucial for the global AI landscape.