A.I. as a widely available commercial product only really entered the public consciousness in November last year, when OpenAI released its hugely popular A.I.-powered ChatGPT. But Khan and the FTC are already seeing issues related to the nascent technology that threaten to undermine fair competition and harm consumers, she wrote in a guest essay published in the New York Times Wednesday.

“While the technology is moving swiftly, we already can see several risks,” Khan wrote, adding that widespread use of A.I. could entrench the status of tech behemoths that already play an outsize role in the economy. 

“A handful of powerful businesses control the necessary raw materials that startups and other companies rely on to develop and deploy A.I. tools. This includes cloud services and computing power, as well as vast stores of data.” 

Left unchecked, leading A.I. companies could dominate the industry much like today’s Google, Meta, and Apple cornered the internet 20 years ago, Khan warned. But if antitrust regulators begin applying pressure now, she argued, we could instead move toward a rosier scenario like what happened when corporate giants such as IBM and AT&T gave up control decades ago—and sparked an economic boom.

Mounting antitrust threats

A.I.’s integration is happening fast, as companies including Google and Microsoft are locked in an arms race to corner the market in a global industry that could be worth nearly $16 trillion by 2030, according to PwC, a consulting firm.

The pace of change has a historical precedent, Khan wrote: The onset of the internet in the 1990s and early 2000s, and especially the emergence of social media and user-generated content online. That technological disruption was an opportunity that today’s tech giants—including Google, Meta, and Apple—seized, and Khan, a longtime advocate of sweeping antitrust reforms, has frequently targeted tech companies for perceived breaches of fair competition rules since becoming FTC chair in 2021. Khan wrote in her guest essay that the online services companies initially promised as free were eventually “monetized through extensive surveillance,” a tradeoff that remains alive to this day.

While the antitrust rules of decades ago were mainly concerned with price fixing, when giants controlled prices while limiting competitors’ say, Khan has been called a proponent of “hipster antitrust,” a school of antitrust that doesn’t just consider price controls as a measure of anti-competitiveness, but also thinks about the structure of markets. So-called market structuralism keeps a small number of large companies at the top through several mechanisms including political influence and swallowing up small potential competitors early on. 

Khan notably wielded this interpretation of antitrust in a campaign against Meta that began last year, where she sued to block the acquisition of a tech startup that Meta saw as a puzzle piece in its larger metaverse vision. 

A similar situation could be playing out with A.I., Khan warned. The space has so far been dominated by a small number of actors, primarily Google and Microsoft, the companies backing the most prolific A.I. labs in the world, DeepMind and OpenAI, respectively. She wrote that because A.I. will likely play a major role in price-setting, collusion between giants “unfairly inflates” costs, and could also lead to “targeted price discrimination.” But she argues that the limited competitive pool also risks hurting the industry’s development.

A lack of government oversight and regulation could lead to A.I. being “trained on huge troves of data in ways that are largely unchecked,” she continued, potentially enabling misinformation and facilitating the work of scammers in ways that threaten “turbocharging fraud.” 

“The history of the growth of technology companies two decades ago serves as a cautionary tale for how we should think about the expansion of generative A.I.,” Khan wrote, comparing the lack of A.I. oversight to the situation that enabled today’s tech giants to dominate the internet. “As the use of A.I. becomes more widespread, public officials have a responsibility to ensure this hard-learned history doesn’t repeat itself.”

But if tech’s consolidation over the past 20 years is a warning about what to avoid, Khan looked even further back to find a blueprint for how to evade predictable pitfalls. Beginning in the 1960s, tough antitrust action coerced large tech companies to disclose the closely guarded inner workings of their success, in a lesson that Khan points to about how to “handle technological disruption for the benefit of all.” 

Cracking down on tech

Khan’s brand of hipster antitrust has plenty of critics, but she wrote in her essay that traditional antitrust action taken decades ago could be a road map for how the government approaches A.I.  

Khan pointed to the case of computing titan IBM, which in 1969 was slapped with an antitrust lawsuit that would eventually lead to a legal battle lasting 13 years. At the time, IBM boasted 70% of the computer market share, and the government sought to split the giant into several smaller companies that could compete with one another. Under mounting government pressure, and as legal fees climbed in a case that involved nearly 1,000 witnesses, IBM unbundled its software from its hardware products, an act some commenters look back on as crucial to shaping the modern software industry, and that Khan said created “trillions of dollars of growth.”

Another example cited by Khan was a 1974 antitrust case against telecommunications giant AT&T, which at the time was the world’s largest privately owned company. A legal settlement in 1982 forced AT&T to divest its 23 local subsidiaries that enabled the company’s monopoly over U.S. telecommunications.

The AT&T breakup led to a 19% increase in telecommunications patenting between 1981 and 1990, while the number of new patents between 1970 and 1981 largely stayed the same or even decreased, according to a study published last year by CEPR, a European economic think tank. The antitrust crackdown “unleashed decades of innovation and spurred the expansion of countless young firms,” Khan wrote.

Lawmakers have largely taken a slow approach to setting the rules around A.I., with some experts saying that the tech is advancing much faster than most policymakers can keep up with. 

Secrets behind companies’ rapid advancements in A.I. have also so far remained protected. Even OpenAI, the startup that was originally founded with the mission of democratizing A.I. for all to benefit, has been largely private about the finer details of its latest systems, notably eliciting the scorn of cofounder Elon Musk, who has since gone on to announce his own rival A.I. startup called TruthGPT.

But the urgency of staying in control of the A.I. industry has not been lost on some government officials. On Thursday, the White House unveiled new measures to ensure A.I. research in the U.S. incorporates ethics and that companies were following best practices to ensure the public equally benefits from the technology. Also on Thursday, President Biden reportedly organized a meeting with CEOs of leading A.I. companies to discuss implementing the measures.

The U.K. announced similar action on Thursday, when the government’s Competition and Markets Authority announced a sweeping review of the country’s A.I. industry, examining whether markets will remain competitive and if the “rapidly scaling technology” threatens to undermine consumer protection laws.