FTC focused on competition and dominance in generative AI space

In her remarks at the summit, Chair Lina Khan stressed that there is “no AI exemption for laws on the books,” and expressed concern about strategies that could “threaten” competition.  Chair Khan also described how market power could lead to violations of consumer protection laws and warned that claims of innovation should not be used as a “cover for breaking the law.”

FTC summit focused on “three layers” of AI: chips and cloud computing, data and models, and consumer applications

Chips and cloud computing

Panelists at the FTC summit discussed concentration in both the cloud computing and chip manufacturing industries.  Panelist Daven Rauchwerk, a technologist and engineer, discussed vertical integration in the chip production market.  According to Rauchwerk, some large tech companies have begun producing their own processor chips required to operate generative AI technologies, making it difficult for new entrants to come to market.  Corey Quinn, Chief Economist at the Duckbill Group, stated that given the high level of concentration in the industry, the largest chip producer has created a supply chain bottleneck with minimal transparency concerning how, and to whom, the chips are being distributed. 

Quinn also described the risk of centralization in the cloud computing industry, which can make it expensive for customers to move data between providers.  He also pointed to an inherent tension that results from the fact that the federal government is “critically dependent” on the three main “hyperscalers” in the cloud computing industry, which he thinks could affect its enforcement priorities.  Tania Van den Brande, Director of Economics at Ofcom, discussed the UK Competition and Markets Authority’s recent cloud services market investigation, during which the agency identified potential harms to competition in the cloud computing industry.  These include barriers to switching such as egress fees that make it expensive to move data between clouds, and discounting structures that create incentives for large customers to maintain all or most of their cloud computing business with one provider.  Ganesh Sitaraman, Professor of Law at Vanderbilt University, highlighted the risk that vertical integration in the cloud computing industry may lead to less innovation in the model or application layers over time, asserting that dominant companies can copy the ideas of smaller startups and integrate them into their own systems. 

Data and models

Panelists also discussed the accumulation of data as a barrier to entry with respect to the development of large language models.  Stephanie Palazzo, a reporter covering AI at The Information, explained that a handful of companies receive large amounts of venture capital funding while other, smaller startups struggle to acquire the capital required to pay for the expensive processor chips and highly skilled talent required to run these models.  Amba Kak, Executive Director of the AI Now Institute, explained that without sufficient funding, smaller companies will not be able to strike deals with third parties for valuable high quality data used to train models, and will be unable to compete with well-funded firms that are able to maximize access to publishers and the media industry and push for exclusivity with respect to data access.  Krisha Cerilli, Deputy Assistant Director of the FTC’s Technology Enforcement Division, expressed concern that access to the large quantities of data required to build large language models and other innovations will steer AI development in a way that concentrates market power even further.

Consumer applications

The third panel addressed consumer protection issues.  Panelist Karen Hao, a journalist and contributing writer for the Atlantic, identified transparency as the biggest risk that AI models pose for consumers, pointing specifically to (1) ambiguous or deceptive marketing, and (2) a lack of clarity for consumers concerning the underlying model embedded in the consumer-facing product with which they are interfacing.  Conrad Kramer, co-founder and CTO of Software Applications Incorporated, discussed how the quality of consumer-facing products depends on what data the underlying model is trained on, and explained that closed-source models do not provide the transparency necessary to allow for informed consumer decision-making.  Andy Hasty, attorney at the FTC, highlighted the risk to consumers of companies marketing their applications with labels like “privacy enhancing” or “AI safety,” when in fact the company’s practices may not align with consumers’ understanding of those terms.

FTC Inquiry into Generative AI Investments and Partnerships

Section 6(b) of the FTC Act authorizes the agency to conduct studies to gain a deeper understanding of market trends and business practices and require the production of information in furtherance of the study.  In its inquiry into generative AI investments and partnerships, the FTC has issued orders to two generative AI companies and three major cloud services providers regarding their recent investments and partnerships.  The FTC has asked for information and documents concerning:

• Strategic rationale behind the investment/partnership;

• Practical implications of a specific partnership or investment, including decisions around new product releases, governance or oversight rights, and the topic of regular meetings;

• Analysis of the transactions’ competitive impact, including information related to market share, competition, competitors, markets, potential for sales growth, or expansion into product or geographic markets;

• Competition for AI inputs and resources, including the competitive dynamics regarding key products and services needed for generative AI; and

• Information provided to any other government entity, including foreign government entities, in connection with any investigation, request for information, or other inquiry related to these topics.

Looking ahead

The concerns expressed by FTC officials and panelists regarding competition in the generative AI arena and adjacent markets are in line with those the agency has previously expressed.³  Competition and consumer protection issues are likely to continue to be top of mind for the FTC as it considers future developments with respect to advanced AI.

Authored by Edith Ramirez, Chuck Loughlin, Logan Breed, and Jill Ottenberg.

References

1 Federal Trade Commission, FTC Tech Summit event page (Jan. 25, 2024) available here.

2 Federal Trade Commission press release, “FTC Launches Inquiry into Generative AI Investments and Partnerships” (Jan. 25, 2024) available here.

3 See Federal Trade Commission Technology Blog, “Generative AI Raises Competition Concerns” (June 29, 2023) available here.