在私募市场，General Catalyst 背书的 Long Lake 以 6.3bn 美元（约 6,300 million）出价收购 Global Business Travel Group（GBTG），这是一家与 American Express 有关联的上市公司，并支付了 65% 的溢价。该溢价尤其引人注意，因为 GBTG 自四年前透过反向合并上市以来已经下跌 50%。Long Lake 的论点是，AI 能借由降低工程师人数与价格，抹掉软体服务过去增长与盈利之间的折冲，从而扩大需求。管理层称，转向 AI agents 可望未来五年每年将毛利率提高最多 2 个百分点，接近 70%；同时 2026 年收入预计略高于 3bn 美元（约 3,000 million），较 2025 年增长 20%，基础自由现金流约为 250mn 美元（约 250 million）。
公开市场投资者仍担心，AI 原生竞争者可能取代像 GBTG 这类既有企业，但 Long Lake 认为可在此之前彻底重组 GBTG。对于 Coinbase 和 Block 这类大而难以被私募收购的上市公司而言，考验更慢：它们必须一季一季证明 AI 的承诺已转化为生产力提升，同时管理市场关于大规模裁员的叙事。私募道路确实能推进更快更大规模的结构重组，但本案也要求高价买断；若并购溢价过高，可能在 AI 效益尚未显现前，先耗掉隐含的 AI 预期红利。
Public and private markets are both betting that AI automation will deliver outsized returns, but reactions to its early workforce moves differ. On Tuesday, Coinbase founder Brian Armstrong announced a 14% workforce cut, and Coinbase shares fell about 2%. In contrast, Block, led by former Twitter boss Jack Dorsey, rose roughly 20% in February after saying it would cut 40% of staff.
In private markets, Long Lake, backed by General Catalyst, bid $6.3bn (about 6,300 million USD) for Global Business Travel Group (GBTG), a listed company linked to American Express, and paid a 65% premium. The premium was especially notable because GBTG had already fallen 50% since going public four years ago through a reverse merger. Long Lake argues AI can erase the previous trade-off between growth and profitability in software services by reducing engineering headcount and prices, thereby expanding demand. Management said shifting to AI agents could raise gross margin by up to 2 percentage points per year for the next five years, toward nearly 70%; at the same time, 2026 revenue is forecast at just over $3bn (about 3,000 million USD), up 20% from 2025, with underlying free cash flow around $250mn (about 250 million USD).
Public-market investors still worry that an AI-native rival could replace legacy players like GBTG, while Long Lake believes it can completely reengineer GBTG before that happens. For listed giants like Coinbase and Block that are unlikely to be private-equity targets, the challenge is slower: they must prove quarter by quarter that AI is delivering productivity gains while managing narratives around mass redundancies. The private route can indeed enable faster, bigger restructuring, but this case required paying up heavily; if the merger premium is too high, a buyer may spend much of the AI dividend before benefits are actually realized.