ABA Panel Reviews Reverse Acquihires in Tech and AI Deals

tang zhengshun via Getty Images

---

---

Summary

On March 26, 2026, panelists in the “Talent or Takeover? New Deals in Tech & AI” session at the 74th Antitrust Spring Meeting of the American Bar Association in Washington, D.C., discussed new deal types in the tech and AI industries, including “reverse acquihires” (or “license and release” deals, which typically involve licensing technology and hiring key employees) and minority share deals. The central question was whether these practices are effectively mergers in disguise designed to circumvent conventional merger review processes. The panel explored the rationale behind these deals, their potential pro- and anticompetitive effects, and how competition authorities worldwide evaluate them.

As Dr. Avigail Kifer, the panel moderator and an economist at Cornerstone, emphasized, AI talent is valuable: training a state-of-the-art model costs hundreds of millions of dollars, so companies compete to hire that talent. Antitrust authorities have raised concerns that AI firms may hoard talent, deny rivals access to talent, and use hiring- or investment-based transactions to sidestep the merger review process.

The panelists agreed that license-and-release deals suit the tech sector because value lies largely in unique, scarce human capital and intellectual property, so acquiring that value provides a real advantage. They discussed potential procompetitive effects, including more closing options for startup founders, increased competition for talent, and stronger innovation incentives, alongside possible anticompetitive effects such as competitor elimination or shelving of innovation.

The panel emphasized that these deals are unique and should be evaluated case-by-case without broad presumptions of harm. They also agreed that agencies worldwide can efficiently review tech deals with existing tools, noting that it is too early to fully assess authorities’ reactions, as few investigations have concluded. Finally, the panelists highlighted that remedies pose challenges since employees cannot be forced to move and licenses are non-exclusive, so authorities prefer preventing potential issues during pre-deal review rather than addressing them afterward.

---

---

---

---

Reverse Acquihire or License and Release?

Leah Brannon from Cleary Gottlieb, explained that the term “reverse acquihire” may not be appropriate. The term “acquihire” emerged roughly twenty years ago to describe transactions in which a larger firm acquires a smaller company primarily to gain access to its employees. More recently, the media have begun labeling certain transactions as “reverse acquihires,” such as the $650 million Microsoft-Inflection deal. In that transaction, Microsoft obtained a non-exclusive license for Inflection’s AI model and hired most of its employees but did not acquire the company itself. These arrangements are neither “reverse” (because the smaller company does not acquire the larger one) nor true acquihires, since no acquisition occurs. Instead, they involve licensing technology and hiring personnel. Thus, “license and release” is a more suitable term for such deals.

According to Leah Brannon, these deals can vary, but they usually involve a non-exclusive license to technology and the release of a fraction of employees from non-compete and similar obligations to their employer. Typically, the company granting the license continues to operate, and the employees who are released from their obligations are free to accept or reject an offer from the licensee. Other than the Microsoft-Inflection deal, such deals include the $330 million Amazon-Adept deal in 2024, the $20 billion Nvidia-Groq deal in 2025, and the $2.7 billion Google-Character AI deal in 2024.

Leah Brannon stated that the reasons for structuring deals as “license and release” are likely simple: companies tend to structure deals based on their objectives. If the buyer wants to hire people but does not want to acquire the licensor’s business, as they would in a merger, a license-and-release deal is a reasonable option. When companies compete to hire talented people, these deals can be pro-competitive in the market for talent by encouraging labor mobility and providing incentives to employees of tech companies to create their own startups, obtain funding, and then consider whether they want to sell their company or license their technology.

U.S. Regulatory Perspectives

On March 25, 2026, the FTC and DOJ announced a request for public comments on premerger reporting requirements, including whether the reporting forms should be updated to address “acquihires,” so-called “reverse acquihires,” certain non-exclusive intellectual property licensing transactions, and other novel deal structures.

Craig Minerva from Axinn and a former lead attorney with DOJ, provided an overview of how U.S. regulators approach these new deals. He highlighted that regulators are still grappling with how to treat these transactions and how they fit into HSR reportability requirements. The regulators’ concerns focus on whether these deals foreclose competition and restrict access to talent, and whether they are an avoidance device to evade formal premerger notification requirements. Craig also quoted Chairman Ferguson, who opined that the increase in these deals is likely related to “excessive enforcement” under the Biden administration and that such enforcement is no longer necessary.

Some of these deals, Craig Minerva pointed out, transfer all or nearly all of a target’s assets to the buyer without involving an acquisition of stock or hard assets. Thus the key regulatory question is whether acquiring human capital and non‑exclusive intellectual property constitutes a transfer of assets under the HSR Act; the current deputy director of the FTC Bureau of Competition thinks it likely does.

Craig Minerva also emphasized that the HSR Act is not the only tool at regulators’ disposal. Other tools include Section 7 of the Clayton Act and Section 2 of the Sherman Act. According to him, minority stakes and other deal types pose fewer issues; the focus is typically on the incentives and ability to influence a competitor’s decision‑making and to obtain competitively sensitive information. Finally, he noted that no U.S. deal investigations have been completed so far.

European Regulatory Perspectives

European authorities have a broad enforcement toolkit and are willing to use it, but they have challenged relatively few recent transactions. Janine Hulsmann of Weil noted that while trends in the EU resemble those in the U.S., there are two key differences: first, these deal structures are unlikely to be driven by antitrust considerations, as EU regulators treat them as potential concentrations and have issued guidance accordingly; and second, designated gatekeepers under the Digital Markets Act (DMA) are required to publicly disclose their transactions. Most transactions disclosed to date are relatively small, and none have been investigated.

Janine Hulsmann additionally opined that in Europe the types of transactions are driven by startup founders’ preferences, since founders are usually the key talent and often have multiple options. Some founders do not want a “reverse acquihire” if they strongly believe in their company and prefer a straight acquisition; others who doubt their startup’s prospects and do not want to disappoint investors may prefer “acquihires” to repay investors and potentially start a new company.

According to Janine Hulsmann, the European Commission has asserted that if a transaction involves the transfer of essentially all the assets necessary to transfer the target’s position in the market, and the agreement causes a structural change in the market, it will be treated as a concentration. German and UK regulators have expressed similar views. However, since the Microsoft-Inflection deal no other similar deal has been investigated, and the Inflection investigation was conducted only by the CMA. Other than that, there have been several minority‑share investigations to date.

Economic Analysis of the Deals

Dr. Lesley Chiou, an economics professor at Occidental College, contrasted the economic analysis of a true merger with that of newer deal forms, noting the key question is whether a transaction generates complementarity or suppression. Complementarity can arise when a larger firm brings resources (e.g., computational power) that a smaller firm’s talented team can use; suppression can occur when a deal eliminates a niche competitor or shelves innovation. The further a transaction is from a traditional merger, the more important allegations around exclusionary effects become—e.g., can the smaller firm continue developing its product after the deal, and is the IP license exclusive.

Competitive effects of these arrangements could be evaluated by comparing the but‑for world to the actual outcome: a startup might have grown into a challenger, or it might use license proceeds to keep developing or pivot to other markets. Firms may have incentives to hoard talent to block rivals, though labor mobility can mitigate that concern. Analysis is also complicated by the absence of existing customers and products in many of these cases.

Recent empirical research suggests these deals can boost innovation and entry; banning all acquisitions can reduce startup formation, while blocking only high‑priced deals may slightly increase entry. These studies report average effects, however, and any particular deal can have the opposite outcome. Professor Chiou also observed that cross‑ownership and similar structures could, in theory, diversify and accelerate innovation via spillovers but might soften competition through common‑ownership effects. Because these innovations are costly and risky, founders weigh exit options: the availability of acquihires, reverse acquihires, and other deal types may raise startup formation by expanding exit paths, but if large firms absorb key talent—or only hire founders—startup hiring costs can rise and incentives for other employees to join startups may fall.

---

---

NOTICE ON AI USE: I used AI (GPT 5 Mini, GPT 5.2) for grammar and stylistic edits. No AI tools were used in the initial drafting of this article. I reviewed the article before submission, and I attest that I agree with its contents. All errors are mine.

---

Endnotes

---

Author

Gleb Domnenko

NERA Economic Consulting

...

View Bio →

---

Author

Gleb Domnenko

NERA Economic Consulting

---

Committees

This content was produced by:

• Economics Committee

Related Content