Competition authorities in the United States, the European Union, and the United Kingdom released a joint statement on competition in generative AI and other AI products, continuing the drumbeat that policymakers need to ensure a level playing field in the next technology wave. The statement identifies three competition risks for AI: the concentrated control of key inputs, the entrenching or extension of existing market power, and arrangements involving key players that could coopt competitive threats.
First, the statement highlights that certain technologies and expertise are critical to foundational AI models, and a small number of firms could exploit these bottlenecks to their advantage, and disadvantage innovation and fair competition.
Second, it highlights that large incumbent digital firms could use their advantages in the digital economy to protect against AI-driven disruption or extend their market power to AI-related markets.
Third, the statement asserts that the major firms involved in partnerships and investments in generative AI could use these arrangements "to undermine or coopt competitive threats and steer market outcomes in their favor at the expense of the public."
The statement also notes that there are other competitive and consumer risks involving AI, including algorithms that enable competitors to fix prices or engage in unfair price discrimination and the risk that firms will "unfairly use consumer data to train their models."
To combat the aforementioned competition risks, the competition authorities set forth several common principles to enable competition: fair dealing, interoperability, and choice. For fair dealing, the statement warns that firms with market power who engage in exclusionary tactics can discourage investment and innovation by other parties. Under interoperability, the competition authorities state that competition in AI is likely to increase if AI products, services, and inputs are interoperable. Finally, concerning choice, the statement notes that scrutinizing how companies use lock-in mechanics as well as investments and partnerships between incumbents and newcomers can foster a competitive process that provides choices among business models and products.
Key Takeaways
The focus on AI is not new but the statement offers three critical takeaways.
- The agencies continue to publicly demonstrate coordination across jurisdictions. For example, in 2021, the U.S. Department of Justice (DOJ), Federal Trade Commission, and the Directorate-General for Competition of the European Commission launched the EU-US Joint Technology Competition Policy Dialogue. The Dialogue includes high-level meetings and regular staff discussion focused on competition issues in technology markets. The Dialogue recently met in April 2024.
- The agencies recognize the evolving nature of AI but do not specify what qualifies as generative AI or other AI markets, ecosystems, or firms. The lack of clarity creates uncertainty concerning what conduct and parties are within the scope of the statement. Companies should consider whether the competition authorities could plausibly consider the markets they operate in as AI markets and if the competition authorities could construe their conduct, inter alia, as attempting to block competitors, control key inputs, or set prices. In the United States, the DOJ has filed statements of interest in cases that allege companies have used algorithmic pricing to engage in price fixing. In Europe, different stakeholders have called for increased antitrust enforcement against alleged anticompetitive conduct in the cloud computing sector on both the EU and Member State level. On June 28, 2024, the French Competition Authority released a report on competition and generative AI and recommended that the EU consider designating companies that host generative AI models as gatekeepers under the Digital Markets Act.
- The statement signals potential scrutiny for mergers and acquisitions, minority investments, and partnerships and agreements involving AI firms, even if the firms involved do not have any horizontal overlap or significant market share or the transaction falls short of a merger. In the UK, for example, the Competition and Markets Authority (CMA) launched a Phase I merger inquiry into Microsoft's hiring of employees of, licensing with, and investment into Inflection AI. The CMA is also investigating Microsoft’s partnership with OpenAI and Amazon’s partnership with Anthropic, though no formal Phase I reviews have yet been launched. Further, in 2024, the FTC launched an inquiry into investments in generative AI companies to understand their impact on competition. The FTC has also launched an informal inquiry into Amazon's agreement with AI start-up Adept to hire some of the start-up’s top executives and license some of its technology. Lastly, while the EU recently concluded that it could not review Microsoft’s investments into OpenAI under its merger rules, it is investigating the partnership to assess whether it is an agreement that restricts competition on account of possible exclusivity provisions. Also, the EU recently started an investigation into Google Gemini’s integration into Samsung handsets.
