Silicon Valley used to reward innovation. Now it strives to sabotage it.

Received wisdom in the tech world is that big, legacy companies are bound by inertia. The more established they are, the more they get set in their ways — and the more vulnerable they are to disruption by a nimble startup. Silicon Valley was founded on the principle that newcomers can move fast and break things — leading to world-transforming innovations.

For the most part, though, that's not how it works anymore. Last year, a couple of economists found that venture-capital-backed startups almost never lead to a new company listing on a public stock exchange. They don't replace the tech giants — they just get bought by the tech giants. That's been true in Silicon Valley for at least a decade. And the vast majority of startups have been acquired by the same five companies: Alphabet, Amazon, Apple, Meta, and Microsoft.

Ever since Joe Biden was elected, the Federal Trade Commission and the Justice Department have been looking into tech mergers and acquisitions for evidence of antitrust behavior. Have the tech giants been illegally short-circuiting competition by buying up their rivals? It's a really good question — but it may be the wrong one to ask. A new paper by two leading scholars suggests that these days, Big Tech doesn't have to resort to buyouts to crush aspiring startups. Instead, they're using their considerable cash and soft power to defang potential rivals from within — a process the scholars call "co-opting disruption."

According to the new paper — by longtime tech observers Mark Lemley at Stanford University and Matt Wansley at Cardozo School of Law — the tech giants are deploying a bunch of sneaky corporate judo moves to undercut their competitors. When an innovative threat to their business model comes along — everything from self-driving cars to virtual reality to artificial intelligence — they make a point of helping the promising little startups. They take a seat on their board of directors. They give them huge infusions of cash and access to selected data. They even lobby the government on their behalf. But all the while, the scholars found, Big Tech is actually using its seemingly benevolent show of support to protect its own interests — subtly steering the startups away from innovation, and into projects that reinforce the status quo. By working from within, they're turning the baying wolves of ambitious startups into harmless, well-trained lap dogs.

This problem isn't academic. Co-opting startups not only quashes competition — it kills off meaningful innovation. Nearly all of the self-driving-car companies have shifted away from their big-vision plans for robot taxis to small-bore stuff like adaptive cruise control. Virtual-reality companies are working on virtual meeting software for the metaverse. AI companies that might have revolutionized drug discovery or engineering optimization are instead fiddling around with search, or customer service, or college-essay writing. With fewer rivals rising to challenge the incumbents, the pace of innovation slows. Some economists even think other startups in the same lanes as the co-opted ones stop trying as hard, for fear of getting bought up and killed off by Big Tech.

This new theory of anticompetitive behavior could be a game changer when it comes to curbing the unprecedented power of the tech giants. It offers regulators a way to crack down on monopoly-minded companies before they try to buy up innovative young guns. "The traditional way of thinking about the threat to innovation from reduced competition has to do with market concentration," Wansley says. "Now we're taking it a little further. We're saying, by the time you get to an acquisition stage, that might be too late."

Lemley and Wansley start their inquiry with a question: Why have two decades elapsed since the rise of the last massive, world-dominating disruptive company in tech? Apple and Microsoft were founded in the mid-1970s; Amazon and Google in the 1990s. Facebook, the baby of the bunch, was founded in 2004. The iPhone came out in 2007. What do we have to show for the past 20 years of venture capital?

The answer appears to be: just another industry that's terrified of competition. At this point, Big Tech looks at promising startups the way evil alien empires in science fiction look at helpless planets. Sometimes they act like the Borg from Star Trek, assimilating the "biological and technological distinctiveness" of competitors to make themselves more badass. Here on Earth, buying up little guys to gain access to their technology or personnel is called an "acqui-hire," like when Facebook bought Oculus and turned it into Quest. And if the smaller company doesn't want to be acqui-hired? Well, resistance is futile.

Or sometimes, like the genocidal, xenophobic Daleks from "Doctor Who," powerful companies just straight-up exterminate their competition. In 2021 a team of researchers from Yale and the London Business School found that somewhere between 5.3% and 7.4% of mergers and acquisitions in the pharmaceutical industry were "killer acquisitions." The smaller company had a drug that might have someday threatened the big guy's category-busting pill, and: pew pew. No more little guy.

In a sense, a dominant company has no choice but to act this way. It can't really innovate anymore — not with thousands of middle managers defending fiefdoms, clients and customers bought into a product and marketing cycle, and hundreds of millions of dollars invested in technical infrastructure. Real disruption would be, well, disruptive. 

"This is not something that people write about or talk about, not something that surfaces regularly," Lemley says. "But it's in the air if you're in Silicon Valley and talking to venture capitalists. It's a problem."

And the bigger the company, the greater the financial risk of any disruption, no matter how small. Like an evil king facing a prophecy that he will one day be overthrown by a firstborn child from the forest realm, the company has to kill all the children, just to be sure.

The thing is, mass murders of forest-realm firstborns tend to attract the attention of regulators. So what if the evil king just adopted all the babies? Raise them rich and spoiled in the castle, distract them with nubile romantic prospects, teach them to exploit the peasants? Between assimilation or assassination lies "a third possibility," Wansley says. "It's not that the acquirer is going to take the startup and totally shut the technology down. It's that they will redirect the assets to something more profitable."

That's what the tech giants are doing. How? First, by relying on Silicon Valley's network of venture capitalists as an early-warning system for would-be disruptors. VCs see trends and startups that might someday pose a threat to the established order — so-called nascent competitors — before anyone else. And there's nothing stopping them. As Lemley observes, it's part of Silicon Valley's "natural information flow" for an investor to tell a Google or a Microsoft about the promising newcomers they're backing. "If I'm a VC, how am I going to get the best deal for my client in this space?" Lemley says. "The best deal might be, sell out to the incumbent."

Once a tech giant gets a tip-off about a startup that might threaten their bottom line, they're able to strategically invest their vast cash reserves in the fledgling firm. That's precisely what's taking place with artificial intelligence: Microsoft owns 48% of OpenAI, the emerging field's undisputed leader. Alphabet and Amazon have put billions of dollars into the AI startup Anthropic. When the two AI companies were founded, Wansley says, both were "worried about what would happen if one of the tech giants started looking in their direction." Now, both are directly influenced by the very companies they set out to disrupt. And since investments from Big Tech often come with a seat on a startup's board of directors, Wansley adds, the tech giants are in an ideal position to "start nudging the company in a direction that's going to be less competitive."

Money isn't the only weapon that Big Tech can deploy to co-opt a potential competitor. Because the tech giants have amassed huge repositories of data on user behavior, they can choose with whom they share that data, and how much. Documents leaked from the discovery phase of a lawsuit against Facebook in 2015 showed that the company doled out data access preferentially, allowing its allies more access than its potential competitors. The data that Big Tech shares — or doesn't share — can play an instrumental role in shaping a startup's work.

Finally, the big companies use their clout on Capitol Hill in an effort to impose stricter regulations on the startups they're ostensibly trying to help. That's why tech giants like Meta testify before congressional hearings and ask for more government oversight for emerging threats like robot cars, or AI. They want to ensure that the rules favor them — and disfavor startups outside their zone of protection. Big Tech may hate regulation, but it doesn't mind using it to regulate any startups that might wind up threatening its dominance in the marketplace.

Co-opting startups is a clever strategy. It's way easier — and less obvious — than buying up a competitor and shutting it down. And if it doesn't work, the nuclear option is always available. "If all those other strategies fail — investments, taking a seat on the board, playing hardball with data networks, and regulation — if none of them prevents a competitor from growing, then the tech companies can buy it off," Wansley says. "It's a little subtle, right?"

It is — but that's what makes it so effective. Straight-up killer acquisitions are relatively easy to prove, says Florian Ederer, an economist at Boston University. "It's much, much harder to prove that a company got taken over and its focus got slightly shifted to not competing quite so intensely." It's even harder to prove that consumers suffered any ill effects. The co-opted companies don't get killed off, and they stay nominally independent. The founders and investors get rich either way.

So how do we know startups are getting co-opted by Big Tech? One proof is that some of Silicon Valley's most powerful players are starting to grouse about it. Elon Musk is suing OpenAI to keep the company from sharing its tech with Microsoft. The venture capitalist Marc Andreessen tweeted last month that Big Tech and the "New Incumbents" in AI are "lobbying as a group with great intensity to establish a government protected cartel." You know things are bad when the biggest kids on the block start whining about being victims.

Wansley and Lemley argue that regulators need to step in — not to restrain competition, but to unleash it. For starters, they say, the feds should strengthen and enforce existing rules against "interlocking directorates," which are supposed to prevent executives from sitting on the boards of their competitors. In the 2000s, the CEO of Google was on Apple's board! These days that kind of thing gives the feds a frowny face; in 2022, the Justice Department pushed seven directors representing venture firms to resign from the boards of five of their direct competitors in space, edtech, and green energy. Microsoft currently enjoys a seat on OpenAI's board of directors. Technically it's only an observer role, without a vote. But, come on.

The primary goal of the scholars I spoke with is broader than any specific regulation: They just want to get government watchdogs to recognize the new and subtle ways that tech giants are shutting down competition and innovation. Sure, people are still founding tech companies, and their startups are still getting funded. But more and more, they're basically serving as farm teams for the majors. "The incumbents have gotten much more attuned to competitive pressures from nascent competitors, and address them more aggressively and earlier," Ederer says.

Regulators, for their part, appear to be catching on. In late January, the Federal Trade Commission ordered three titans of tech — Amazon, Microsoft, and Alphabet — to open the books on their investments in artificial intelligence. OpenAI and Anthropic received the same letter: Send over every scrap of documentation on the financial ties between them and the massive incumbents that are funding them. 

That's a promising start. If technology is going to do what it always has — solve big problems, build value, create jobs and opportunity — Silicon Valley needs as much competition as possible. "We used to have these cycles of competitive disruption," Lemley says. "Microsoft runs the world, and the internet comes up, and they missed it. IBM runs the world, and the Dells of the world come up, and they're no longer in charge. That's what we'd like to get back." Startups once unlocked a golden age of tech innovation. Perhaps, if we can find a way to keep them from being co-opted, they can do it again.

Adam Rogers is a senior correspondent at Business Insider.