top of page

Competition Regulation and the A.I. Arms Race

What is Competition Regulation and Why is Important?


Competition law creates checks for practices which restrict competition “to the detriment of consumers.” Creating fair competition encourages innovation, improves product and service quality, and keeps the prices down. Whilst specific regulation varies by jurisdiction, the checks in place in each competition regime prevent anticompetitive behaviours including but not limited to abusing a dominant market position, creating long term exclusive contracts with suppliers or customers, and restricting the price other businesses can sell your product for. Cartel activity is the most serious anticompetitive behaviour which entails businesses agreeing not to compete with one another using methods such as price fixing, bid rigging, sharing markets or customers, and sharing commercially sensitive information. These illegal agreements can be formal or informal and there are heavy penalties for infringements. Additionally, under Competition Law mergers and acquisitions (M&A) may be investigated. If it is suspected that the transaction will have an anticompetitive effect the acquisition or merger will not take place until the competition authority - such as the European Commission (EC) in Europe- has subjected the acquisition to various tests and given its approval. 


Global Competition Regulation Trends in the U.K., E.U. & U.S.


Each jurisdiction affords differing amounts of power to their regulators, and the way in which countries are promoting competition does diverge. That said globally, M&A antitrust agencies are increasing merger control and there appears to be more of a focus on vertical mergers, which were previously considered pro-competitive. In 2023 the Federal Trade Commission (FTC) and Department of Justice (DoJ) formalised their altered merger guidelines, the E.U.’s new Digital Markets Act (DMA) came into effect and the U.K. introduced the Digital Markets, Competition and Consumers Bill (DMCC). Antitrust in 2023 was marked by policy reform as more pressure was placed on authorities from elected officials following years of perceived “under-enforcement." 

The Biden administration in particular seems to have taken a ““big is bad” approach to antitrust” not only in technology but also other sectors, such as healthcare. In 2021 Biden issued an Executive Order (EO) which directed the FTC and DoJ to focus on acquisitions made by large tech companies. Biden’s focus on merger law diverts from previous years where it was seen as “less concerning.”


Whilst prima facie the Biden administration’s focus on merger law, and specifically the technology sector, seems excessive- this is not unique to any one jurisdiction. Other countries have similarly or are in the process of introducing new digital markets legislation, such as Japan and the U.K. 


That said, some jurisdictions have a reputation for being more stringent than others. Following Brexit, the U.K.'s Competition and Market Authority (CMA) gained a reputation for being the “most interventionist antitrust agency in the world.” Further to this, the CMA’s authority, which was already bolstered following Brexit, will be increased even more when the DMCC comes into effect in the Autumn of this year. This is an interesting move from the U.K. government as the increased regulation of Big Tech undermines Rishi Sunak’s efforts to lure Big Tech and A.I. development to the U.K.. There is little doubt that the U.K. government is performing a “delicate balancing act” between economic gain and preventing under-enforcement. One such example of this balancing act was written about this December in the Technology Review. The article claims that Microsoft’s £2.5 billion investment in cloud infrastructure in the U.K. was a “clear attempt to blunt an investigation into the cloud market by the U.K.'s competition regulator.” 


Although this is a substantial claim to make, Big Tech is certainly trying to damper the level of control that the CMA has over their business. This can be seen by the multiple appeals made about the new powers of the CMA. Whilst the Big Tech firms did win many concessions on appeal in November 2023, the new powers of the CMA (to conduct judicial review) were retained. This has remained the case despite protests that the judicial review afforded the CMA, as a body, far too much power and that an appellate system ought to be created.


Although the U.K. is attempting to create a pro-competitive environment, some critics wonder whether the increasingly stringent regulations will undermine the jurisdiction’s attempts to ingratiate itself with Big Tech, and as such undermine Sunak’s ambition to create a prominent A.I. sector there. It certainly has had an effect, previously, on the U.K.'s relationship with Meta, as it emerged that Mark Zuckerberg threatened to pull investment from the country if regulation on Silicon Valley companies was not softened, during the CMA’s investigation into digital advertising in the U.K.. Further to this, in early 2023 Microsoft President Brad Smith also reacted bitterly to the CMA’s initial decision to block Microsoft's bid for Activision, stating that the regulator's move "had shaken confidence" in the U.K. as a destination for tech businesses. Big Tech firms do seem to be taking aggressive stances in response to regulatory scrutiny, as seen by OpenAI’s threat to leave the E.U. after the A.I. Act. 


However, recent events have demonstrated the pull which Big Tech has in the A.I. sector at present, Technology Review writing that ‘until late November, when the epic saga of OpenAI’s board breakdown unfolded, the casual observer could be forgiven for assuming that the industry around generative A.I. was a vibrant competitive ecosystem. But this is not the case—nor has it ever been.’ Microsoft’s involvement in the re-hiring of Sam Altman to OpenAI despite holding only a nonvoting share on OpenAI’s board demonstrates, as is argued in the article, that some form of regulatory scrutiny is needed to temper the level of control Big Tech has in generative A.I. 


The concentrated power which these large companies have is not solely problematic for the tech sector, but also can cause issues for democracy and collective agency. Tech executives, among others in leading positions, have voiced this concern about the development of A.I. being entirely in the hands of too few organisations. Meredith Whittaker, president of encrypted messaging app Signal, said in an interview to CNBC that society ought to be concerned about the “handful of corporations driven by profit and shareholder returns making such socially consequential decisions.” 


Cloud Computing


The involvement of Big Tech in generative A.I. issues stems not only from the initial financing that tech giants provide which A.I. startups need, but also the vast computational power required to run a large language model (LLMs). Such computational power can easily drain the financial resources of a generative A.I. startup. Cloud computing “offers seamless access to servers, networks, storage, development tools and applications via the internet” and is an essential business strategy and a “prerequisite to advancing A.I.” Cloud computing offers an almost endless pool of resources with which to train LLMs, and most of this the global cloud computing market is reportedly controlled by large tech companies such as Amazon and Google, Amazon supposedly owns around a third.


The cloud-computing infrastructure is what allows the link to form from tech giants to generative A.I., causing an extension of the market dominance those tech giants already possess. In April 2023, April, researchers at the AI Now Institute published a report stating that A.I. development was "foundationally reliant" on the resources of tech giants. 


There have also been concerns about the so-called “round tripping” which tech giants are also performing where the money produced by investment into the generative A.I. startups, which were lured to the tech giants by their cloud infrastructure, is then re-invested into the cloud infrastructure. This is also just the beginning, as the biggest cloud computing giants (Amazon, Microsoft and Google) not only announced third quarter annual growth of 28% to 42%, but are also all reportedly upping spending. The arms battle to dominate the cloud market and the generative A.I. space has evidently begun.


Big Tech Giants & the FTC


This arms battle has been noticed by the various regulatory bodies globally and attempts to block acquisitions by tech giants have increased, although the results have not proven successful thus far. 


Recently the FTC unsuccessfully attempted to block Meta Platforms Inc.’s acquisition of Within Unlimited Inc., which is a developer of virtual reality (VR) fitness applications, although this case has potentially positive ramifications for the concept of actual potential competition in U.S. law. Actual potential competition was established as a viable theory of harm in the 1960s and 70s, but has never been successfully brought in court. Actual potential competition is the concept that firms with no current share of the relevant market could still be an “important competitive force.” The Act applies to mergers which have a probable anticompetitive effect, and it has to be shown that the entrant was likely to enter the market. In the Meta case, the FTC were consoled by the fact that the court acknowledged actual potential competition as a viable theory of harm, but the ‘evidentiary bar’ was set so high that it seems the government will have difficulty continuing to advance the theory.


More recently, and perhaps with a higher chance of success, the FTC voted unanimously to investigate tech giants and the deals they have formed with generative A.I. companies, including Microsoft, OpenAI, Amazon, Alphabet and Anthropic. Lina Khan, FTC Chair, said that they investigation would look into whether the tech giants are “distorting innovation and undermining fair competition.” In response to the compulsory order, Google took the opportunity to dig at the relationship between Microsoft and Open AI in a statement which read: “we hope the FTC’s study will shine a bright light on companies that don’t offer the openness of Google Cloud or have a long history of locking-in customers – and who are bringing that same approach to A.I. services.”


I am sure I speak for many others when I say that I am extremely interested to see how the A.I. arms race plays out not only between the tech giants, but also the other involved parties- the regulatory bodies.

Comments


bottom of page