Competition in the AI space has reached increasing highs of late, leading to record levels of investment and novelly-structured deals that depart from the traditional startup lifecycle and may impact the broader ecosystem. As players of all sizes are looking to build and scale quickly, and regulators have approached traditional acquisitions with heightened skepticism, a new type of unorthodox deal has emerged. Leading companies are paying billions to hire startup leadership teams and license their tech rather than acquire the company outright, which can carry downstream impacts for the flow of capital and talent through the startup ecosystem.
Broadly, there are three types of startup exits: going public, being acquired, or failing. Exits are a critical moment in the lifecycle of a startup—after an exit the company is likely to no longer be considered a startup: it will either be a public company, be part of the acquiring firm, or cease to exist. In successful, profitable exits, founders are rewarded for the blood, sweat, and tears they have put into building the company, and investors earn a return. Acquisitions are the overwhelming majority—or sometimes the only—type of successful exits in startup ecosystems across the country. These acquisitions promote the building of knowledge, recycling of talent, and flow of capital through the ecosystem. Each of those components are key to building new startups and stimulating the investment needed to grow them to scale.
Increasingly—especially during the Biden Administration—antitrust enforcement agencies have viewed acquisition activity skeptically. During the Biden years, the agencies rewrote the merger guidelines, jettisoning the commonsense consumer welfare standard and guidance that had been widely accepted by courts. They remade the premerger notification process, increasing the cost and time to prepare merger filings, amounting to an additional tax on merger activity. They brought challenges to deals involving startups with novel pleadings ultimately rejected in court. They opposed other deals involving U.S. tech companies, sometimes seeing them abandoned or quashed with the help of foreign regulators—even though the failed deals led Americans to lose their jobs, destroyed billions in startup value, and reduced startup investment.
This increased skepticism and uncertainty around acquisitions—especially for large deals—changes the calculus for startups when it comes to exit options. If founders are unsure if their proposed transaction will actually close, to avoid the negative consequences of a failed deal, they may choose to pursue other options like entering a licensing agreement for their company’s tech and taking a job at a large firm.
This is becoming more popular amid pitched competition at AI’s frontier, where top tech teams in the AI space are in high demand and many startups are scaling fast to large exits. Some of those exits have been traditional acquisitions, but others have taken on more novel and unorthodox licensing structures. For example, Microsoft paid a $620 million licensing fee to InflectionAI for the company’s tech and to hire the founding team and nearly all its employees in March 2024. Google, Meta, and Amazon have each since done deals with a similar structure, some worth tens of billions. The deals underscore the competition for talent, the value of AI innovation, and the speed of AI development. And they reflect the changed incentives for founders given the environment of hostility around acquisitions.
It is—and should be—startups’ prerogative to decide what exit to pursue as right for them, but these novel deals do upend the traditional archetype of a successful exit and who “wins” from the exit. A significant portion of employee compensation at early-stage startups tends to be in the form of equity in the company, because startups do not have resources to pay large salaries. The promise of a future payoff arising from an exit is key to recruiting key early talent who could instead take higher paying, less risky positions at more established companies. Moreover, individuals and organizations invest in startups with the expectation of a return. Public reporting about the deals that have followed this pattern so far has indicated that investors have been made whole, but maybe not to the same magnitude of other potential exit paths. And whether and to what extent employees win from these novel transactions has varied—some have led to the hire of all of the startup’s employees, others have only involved founders and more senior employees. That could alter what secondary effects arise from exits and could undermine the incentives for joining a more risky startup.
These outcomes that depart from a more traditional acquisition raises important questions about whether and to what extent the beneficial consequences to the broader innovation ecosystem occur, namely the transfer and circulation of knowledge, capital, and talent. Prototypically, exits are transformational for (even junior) employees who often use the resources from the exit to start their own companies and invest in other startups. Capital returned to investors is reinvested in new startups, and any dampened returns will show up in reduced capital formation.
Where these outcomes might be less beneficial than the status quo demonstrates the importance of understanding that incentives set by policymakers matter—especially when those policies are nominally designed to promote competition. Startups need competition policy to work with the elements that drive virtuous cycles of growth, not against them through restrictive policy choices.
Engine is a non-profit technology policy, research, and advocacy organization that bridges the gap between policymakers and startups. Engine works with government and a community of thousands of high-technology, growth-oriented startups across the nation to support the development of technology entrepreneurship through economic research, policy analysis, and advocacy on local and national issues.
