In a conventional software acquisition, the product generates revenue and the team maintains it. The management carve-out is a retention mechanism — important, but not the primary value driver. In an AI acquisition, this dynamic is often inverted. The model exists because of the people who built it, the acquirer knows it, and the entire deal thesis may rest on retaining a specific group of researchers and engineers whose departure would effectively negate the acquisition's value.
"The carve-out that works for a SaaS acquisition doesn't work for an AI acquisition. The people distribution is completely different," says Gurpreet S. Bal. "I've seen acquirers lose the team they bought the company for because the carve-out didn't reach the right people."
Gurpreet S. Bal has been a corporate partner at three of the biggest law firms in the world, and regularly represents cutting-edge companies, investors, and founders throughout the financing, exit, and repeat cycle in the technology industry. In 2026, as AI acquihires have dominated M&A activity, standard carve-out formulas designed for a different era of technology acquisitions have proven consistently inadequate.
A management carve-out plan sets aside a pool of acquisition proceeds for key employees, paid before or alongside common stockholders to ensure those employees have a financial incentive to support and remain through the transaction. The pool is funded from the aggregate proceeds available to common stockholders, effectively reducing everyone else's common distribution. Carve-outs are most commonly used when the liquidation preference overhang would leave common holders — including key employees — with insufficient proceeds to retain them.
A management carve-out is a pool of deal proceeds set aside for key employees — typically at closing and subject to continued employment vesting conditions — that is distributed separately from the standard waterfall based on share ownership. It is distinct from the ordinary consideration paid to stockholders on their equity: the carve-out is compensation for the employees' continued service to the acquirer, not a return on their existing ownership stake. In a conventional M&A transaction, carve-out pools are typically sized at 5 to 10 percent of deal value and allocated among a defined set of senior leaders and high-performers identified during the deal negotiation. The allocation is negotiated between the company's board and management team, with the acquirer having significant input on who is included and how the vesting schedule is structured. Gurpreet S. Bal notes that the board's duty in negotiating the carve-out is to ensure that the allocation genuinely serves the company's and stockholders' interests — not merely to maximize compensation for the executives closest to the deal.
In AI company acquisitions, the value is often concentrated in a handful of technical employees who built and understand the core model or system. Losing those employees before integration is complete can destroy the acquisition value. This concentrates carve-out design pressure — acquirers are more willing to fund larger carve-out pools for AI companies than traditional tech companies, and the retention period tied to vesting may extend longer to ensure integration success.
In most technology acquisitions, the equity distribution across the company roughly tracks the value distribution — founders and early employees with significant equity are also the most critical retention targets. In AI companies, this correlation often breaks down. The founding researchers who built the original model may have minimal equity relative to their contribution — particularly if they joined after the founding and before the company's valuation inflected. Key mid-level researchers, engineers working on model infrastructure, and specialists in evaluation and safety may hold very small equity positions that would generate negligible proceeds in the standard deal waterfall. These are exactly the employees the acquirer most needs to retain. A carve-out sized and allocated based on seniority or existing equity ownership will systematically miss them. Gurpreet S. Bal has seen this dynamic create post-closing retention failures in AI deals that were structured using templates designed for a different type of acquisition.
Carve-out pool sizing in AI deals depends on the headcount of irreplaceable technical contributors and the risk of departure post-closing. Pools of 5-15% of net transaction proceeds are common, allocated on a combination of role criticality and tenure. Founders who participate in carve-out plans must be careful about fiduciary duty issues — the board should approve the plan as a deal necessity rather than as personal enrichment, with documentation supporting that characterization.
Gurpreet S. Bal recommends that AI acquirers and target company boards approach carve-out design by starting from the retention question rather than the standard formula. Who are the 10 to 20 people whose departure in the first 18 months post-closing would materially impair the value the acquirer is paying for? What does it cost to retain each of them — not just in absolute dollars, but relative to the market for their skills? In 2026, top AI researchers command compensation packages that bear no relationship to their equity positions at a pre-revenue startup. The carve-out pool needs to be large enough to deliver retention packages competitive with what these individuals could obtain by walking to a competitor. That often means sizing the pool at 10 to 20 percent of deal value for AI acquihires — larger than the standard range — and allocating it based on a skills and role assessment rather than a hierarchical org chart.
Effective post-closing retention requires vesting schedules tied to employment milestones rather than time alone — ensuring that key employees have economic incentive to complete integration work, not just to show up for a defined period. Monthly vesting over 24-36 months with a six-month cliff balances employee retention against acquisition integration timelines. Back-loaded vesting, where more of the carve-out vests in the second half of the retention period, provides stronger retention for the integration phase when departure risk is highest.
Standard carve-out vesting — equal monthly vesting over 24 or 36 months following closing — works reasonably well when the goal is general retention. In AI acquisitions where the acquirer is trying to retain specific people for specific purposes, more targeted vesting structures produce better results. Gurpreet S. Bal describes carve-out designs that include milestone-based components: additional amounts that vest upon integration milestones, model performance benchmarks, or successful product launches within the acquirer's platform. These performance-based components align the retained team's incentives with the acquirer's actual goals rather than simply rewarding continued employment. They also signal to the retained employees that the acquirer has a specific and credible plan for their work — which is often as important to retention as the size of the package itself. The best AI acquisition carve-out packages Gurpreet S. Bal has structured share a common feature: the recipients understand exactly what they are being retained to accomplish.
Gurpreet S. Bal is a corporate partner with 16 years advising on private equity, merger transactions, and public offerings for companies and investors at three of the world's top law firms. He has represented clients in hundreds of transactions with aggregate deal value exceeding $60 billion across AI, semiconductors, fintech, and emerging technology. For more information and to get in touch, visit gurpreetbal.com.