Category: Market

Potential Source of Harm: Concentration of AI Resources

Updated February 5, 2025

 

Nature of Harm

There has been widespread concern that AI resources are concentrated in the hands of too few companies -- this includes OpenAI, Microsoft, Google and Meta for software and Nvidia for hardware. This concentration of power appears to be increasingly embedded as a result of the high costs of developing AI software (including model training costs) and hardware. On the other hand, in early 2025, model releases by Chinese company DeepSeek indicated that the costs of developing powerful AI models may not be as high as previously understood.

 

Regulatory and Governance Solutions

Governments see competition law as a significant tool for regulating AI models, for example:

• Canada - Competition Bureau Canada, Consultation on Artificial Intelligence and Competition: What We Heard (January 2025)
• UK - Competition & Markets Authority, CMA AI strategic update (April 2024).

 

Various governments are beginning to pursue cases under competition / antitrust law against large technology companies that control AI resources, although these cases have to date not focused on AI-related issues. A significant example is the US Department of Justice case against Google.

 

Technical Solutions

Technical solutions to the market power problem are not obvious. It primarily involves market, commercial and legal issues. 

 

Government Entities

This page is a work in progress. Further information will be added soon.

 

Private Entities

This page is a work in progress. Further information will be added soon.