US, UK, and EU share views on competition in AI development

4 months ago 21

The UK, US, and EU competition authorities specified common fair competition principles on foundation artificial intelligence (AI) models like ChatGPT but did not commit to joint action, according to a joint statement on Tuesday (23 July).

Foundation models, akin to general-purpose AI, are considered some of the most powerful ones due to the vast number of applications they have. They underpin generative AI, like ChatGPT, which can create text, images, and video, based on prompts.

Competition authorities for the three jurisdictions have been looking into how partnerships between the developers of these AI models, like OpenAI, and Big Tech companies, affect market competition.

“Our legal powers and jurisdictional contexts differ, and ultimately, our decisions will always remain sovereign and independent,” said the Tuesday statement, signed by the EU Commission, the UK’s Competition Markets Authority (CMA), and the US Federal Trade Commission (FTC) and the Department of Justice.

But due to the international nature of these problems, it continued, “we are working to share an understanding of the issues as appropriate and are committed to using our respective powers where appropriate”.

In the statement, the competition authorities pointed to three risks: concentrated control of key inputs like data or cutting-edge AI chips, entrenching or extending market power in AI-related markets, and partnerships among key players.

Cutting-edge AI chips and models are owned, but not openly shared, by Big Tech companies.

To address the risks, the competition authorities will focus on ensuring consumer choice, interoperability of different services, and preventing exclusionary tactics.

The risks are only slightly revised from three key risks specified in a statement from the UK Competition and Markets Authority (CMA) in April.

The CMA has been looking into various partnerships between Big Tech companies and foundation model developers, such as Amazon and Anthropic. Tuesday’s statement “does not alter its merger review process in relation to its AI partnership cases”, which also include Microsoft and OpenAI’s partnership.

The EU executive has reportedly dropped a merger investigation into Microsoft’s investment and partnership with OpenAI but is still examining an anticompetition angle.

The FTC launched its own investigation into such partnerships in January.

[Edited by Eliza Gkritsi/Zoran Radosavljevic]