Proposed antitrust remedies for Google ignore the impact of AI on internet search

Closing arguments in the remedies phase of the antitrust case against Google’s search engine ended on May 30, with a decision expected from U.S. District Judge Amit Mehta in August. The remedies sought by the Department of Justice (DOJ) include a forced sale of the Chrome internet browser, bans on default contracts with other browsers, and mandated sharing of data with competitors.  

In the five years since the case was filed, however, the market for internet search has been upended by artificial intelligence (AI). In a revitalized market undergoing rapid innovation and unexpected change, the remedies proposed by the DOJ will not open competition as intended and will become less relevant with time in the wider market. They will simply harm Google and its ability to compete at the intersection of search and AI. 

In August 2024, Mehta found Google liable in the case, a verdict Google still plans to appeal. The judge recognized the role that simple product quality has played in Google’s dominance of internet search, but ruled that market conduct, such as contracts for default status that paid Apple and Firefox billions of dollars, served to maintain illegal monopoly power. 

Outside the courtroom, Google is in what many consider the competitive fight of its life. Advances in generative AI technology and public adoption of AI-powered assistants (“chatbots”) mean the alleged monopolist is facing a reality few foresaw in 2020 when the DOJ filed suit. On May 7, the company’s stock took a hit after news that AI chatbots had already made a dent in overall Google searches. Meanwhile, Google rolled out major new AI offerings, signaling its intent not to cling to older search products suddenly at risk. 

Proponents of breaking up big tech have spent most of the past decade treating companies like Google as having reached an endpoint in their evolution—its dominance of internet search, not to mention the centrality of internet search in our lives, frozen in place unless the government intervenes. But today’s seemingly dominant firms, especially in the tech sector, often face new competitive realities tomorrow.  This process, termed disruption or creative destruction, is a major source of competition over time that antitrust authorities frequently ignore. 

People now use generative AI assistants (“chatbots”) to answer many of the questions that previously required an internet search. Much of the necessary innovation in LLMs, and especially widespread awareness in the market, has taken place while Google has been an antitrust defendant. In 2020, when the DOJ filed suit against Google for monopolizing internet search, few expected rapid disruption at the hands of AI. 

Mehta’s fall 2024 decision finding Google liable in the case and triggering the remedies phase contained a lengthy analysis of generative AI that would have been unrecognizable to anyone at the trial’s start. Some of that reasoning has already been called into question by even more recent market events. Last fall, the judge ruled that AI chatbots were outside the relevant market for internet search. Fast forward to May 7, when Apple CEO Tim Cook testified in the remedies phase that losses to AI chatbots caused Google searches in Apple’s Safari browser to decline during March and April for the first time in over 20 years. Google stock fell 7 percent on the news. 

Internet search and generative AI, as currently used in chatbots, are neither competitors nor complementary products. This lack of clarity is driven in part by remaining uncertainty around what kind of products AI chatbots are. Current AI chatbots access enormous amounts of information but don’t update it in real time, an example of the innovative process still underway. Rather than seeking to suppress new technology to keep people using its search products, Google is leaning into developing AI products.  

There is some irony in the timing of these events, happening opposite a trial that was launched because of Google’s supposed insurmountable dominance of a market that some expected to remain unchanged. The DOJ has pivoted in the rationale it presents for why Google should sell its Chrome browser, along with other major restrictions going forward. Its arguments in May centered on why Google must be stopped (by the government) from parlaying its dominance of search into dominance of AI. 

The DOJ believes that because internet search has the real-time capability mentioned above that AI chatbots today lack, it will become a choke point where Google can flex its muscles either to push its own products or make others like ChatGPT pay up. The logic is vague and not fully developed, because everything about the rapidly evolving AI landscape is vague and not fully developed. 

Courts are slow by design, and the interventions they can order from the top down are inherently blunt instruments. Since the DOJ rolled out its proposed remedies, many have rightly protested that the punishment does not fit the alleged crime of monopolizing internet search. Unlike the steel and telephone breakups of the past, it would not be clear along what lines to split the tech giant even under the most draconian penalties. 

It is absurd to think that the same list of brute-force remedies, still subject to years of appeal and implementation, will have anything resembling the outcomes the DOJ now says they will on AI markets. The current pace of innovation makes it impossible to say what the impacts of these remedies would be on AI markets years from now. 

The DOJ’s remedies also risk negative unintended consequences. An ecosystem of developers and content creators, most of them very small, has grown from the bottom up around Google. Divesting Chrome and severely limiting other forms of contracting could limit this ecosystem’s unique capacity to innovate at the intersection of AI and search.  

Google I/O, an annual event attended by thousands and livestreamed by more, also took place in May. The event is where Google traditionally announces new products as well as initiatives of interest to its developer and creator ecosystem. This year’s installment, unsurprisingly, was dominated by products and initiatives using the company’s own AI offerings.  

Google’s Gemini currently ranks third among AI chatbots. No longer a nimble startup, the company will likely play a different role in this round of innovation than it did at the start of the century. However, Google can bring thousands of software developers, app makers, and others in the industry together not just to advertise its own products, but to benefit from the many connections in its ecosystem.  

Nobody could have predicted that the disruption of internet search as we know it would come to a head just as the DOJ delivered closing arguments for its plan to curtail Google’s search monopoly. The DOJ’s pivot from opening the search market to protecting competition in AI should call its credibility into question. The main effect of its proposed remedies will be to harm Google and its future operations. Innovation and competition will continue to create new market realities faster than any court-ordered intervention. 

The post Proposed antitrust remedies for Google ignore the impact of AI on internet search appeared first on Reason Foundation.

Source: https://reason.org/commentary/proposed-antitrust-remedies-for-google-ignore-the-impact-of-ai-on-internet-search/