FTC Warns on Big Tech Partnerships with AI Startups

The U.S. Federal Trade Commission on Friday issued a report outlining concerns with major Big Tech and AI startup partnerships. Key among them, of course, is the Microsoft and OpenAI partnership that kicked off the current AI boom. But Amazon, Google, and Anthropic are each engaged with each other in troubling ways, too, the agency says.

“As companies rapidly deploy generative AI technologies, enforcers and policymakers must stay vigilant to guard against business strategies that undermine open markets, opportunity, and innovation,” FTC Chair Lina Khan said. “The FTC’s report sheds light on how partnerships by big tech firms can create lock-in, deprive start-ups of key AI inputs, and reveal sensitive information that can undermine fair competition.”

This all feels rather obvious: The most likely outcome of this era is that the biggest cloud superpowers–Amazon, Google, and Microsoft–expand their market power thanks to the complexity and expense of deploying generative AI at scale. The only real questions are whether the market power will shift somewhat between them, and what regulators will do to stand in their way.

In any event, the FTC’s 43-page report is the result of several months of investigation into three major Big Tech/AI partnerships: Microsoft and OpenAI, Amazon and Anthropic, and Google and Anthropic. It doesn’t represent a legal move of any, or even the start of a broader set of investigations. Instead, the report seeks to contribute to the public’s understanding of what is a rapidly evolving generative “AI landscape.” And it highlights key concerns for the future. Which, again, are patently obvious.

But here they are. The FTC believes that these partnerships could be used to limit computing resources and engineering talent to non-partners, increase contractual and technical switching costs for developers and customers, and curb the spread of technical and business information related to “generative AI models, AI development methods, confidential chip co-design, partner finances, and customer usage and revenue numbers.”

Of the three partnerships studied, Microsoft/OpenAI is by far the biggest: The FTC values Microsoft’s investment at $13.75 billion, while Amazon/Anthropic is at $8 billion and Google/Anthropic is at $2.55 billion. Of course, these numbers don’t include the vast sums of money each is paying to build out their respective AI infrastructures and direct customer offerings. Microsoft, for example, will spend at least $80 billion on its AI infrastructure in this fiscal year alone. One of the ways the rich get richer is to invest in ways smaller firms cannot.

The report offers a reasonably readable summary of generative AI and how it’s evolving, and rapidly, and it’s worth recommending just for that. But in addition to “protecting and educating consumers,” the FTC’s primary aim is to “promote competition.” And Amazon, Google, and Microsoft will spend whatever is necessary to lock out potential customers through acquisitions, partnerships, and infrastructure, its early take on that issue is perhaps the most interesting aspect of the report. Which, incidentally, is highly redacted because of the private business information the original contains.

Looking at just one aspect of the Microsoft and OpenAI partnership, the FTC notes that the latter firm must use Microsoft Azure as its “exclusive cloud provider” under the terms of its partnership, which started in 2019 and was “reaffirmed” in 2023. Both companies claimed to the FTC that this legal agreement was separate from Microsoft’s investment. But multiple sources told it that was not the case. Obviously, the exclusivity is tied to Microsoft’s massive investments.

Amazon, Google, and Microsoft all communicated various capacity issues related to AI. But their respective spending shuts out most of the market. For example, very few companies outside of the big three can gain access to the GPU resources that are necessary to power today’s cloud AI workloads. And even when these companies do get the hardware they need, they tend to funnel those resources to their own needs, not to those of customers with their own products and services.

But there are no real conclusions.

“Commission leadership has highlighted the great promise of AI to benefit consumers and make American businesses more effective, while recognizing potential risks to competition and consumers, such as ‘the expanding adoption of A.I. further locking in the market dominance of large incumbent technology firms’,” the report explains. The partnerships it highlights are not “a new phenomenon,” it says, but generative AI has lead to a resurgence in this type of partnership, and they merit “merger scrutiny.” It is “important that the FTC and the public understand the terms of the partnerships and their impacts on competition and consumers.”

Sadly, I think we do understand it. And that it’s highly unlikely the U.S. government will do anything meaningful to curb the behavior of Big Tech firms–all U.S.-based–anytime soon.