This Chinese Tech Giant Quietly Cut Its Buyback 80% to Fund AI

Alibaba Group Holding Ltd. cut its share repurchases by roughly 80% in the June 2026 quarter, buying back 13.4 million ordinary shares (1.7 million ADSs) for US$162 million versus 56 million ordinary shares (7 million ADSs) for US$815 million a year earlier, to redirect cash into AI infrastructure. Capital expenditures rose 75% year over year to RMB 67,678 million, while free cash flow fell to negative RMB 44,670 million and the new AI Labs and Applications segment posted an adjusted EBITA loss of RMB 13,861 million. CEO Eddie Wu said on the August 20 call that "AI has become Alibaba's most certain growth engine," and CFO Toby Xu argued the investment has "a clear path to attractive ROIC.

Alibaba NYSE:BABA https://247wallst.com/companies/BABA/ | BABA Price Prediction https://247wallst.com/companies/baba/price-prediction just made its capital allocation priorities unmistakable. In the June 2026 quarter, the company repurchased 13.4 million ordinary shares approximately 1.7 million ADSs for US$162 million. A year earlier, in the same fiscal quarter, it bought back 56 million ordinary shares 7 million ADSs for US$815 million. That is roughly an 80% cut in ADS repurchases at a company that still had US$19.3 billion of authorization remaining as of June 30, 2025. The cash was rerouted into silicon and concrete. Where the Buyback Money Went Capital expenditures rose 75% year over year to RMB 67,678 million for AI infrastructure https://247wallst.com/investing/2026/07/10/top-wall-street-analyst-says-ai-spending-is-delivering-real-returns-when-we-build-a-data-center-its-already-pre-sold/ . Free cash flow deteriorated to negative RMB 44,670 million from negative RMB 18,815 million a year earlier. The newly disclosed AI Labs and Applications segment posted an adjusted EBITA loss of RMB 13,861 million, up from RMB 3,224 million. On top of that, the quarter absorbed a EUR 550 million European Commission fine and RMB 4,458 million goodwill impairment. Internal cash is not covering the buildout alone. During fiscal 2026, Alibaba raised approximately US$3.2 billion in convertible notes and HK$12 billion in exchangeable bonds to fund cloud and international commerce, and total debt to adjusted EBITDA doubled to 2.29x. Full-year FY26 repurchases came in at just US$1.046 billion, a fraction of prior years. Management Frames It as an ROIC Bet CEO Eddie Wu was direct about the shift. “AI has become Alibaba’s most certain growth engine,” he told analysts on the August 20 call. CFO Toby Xu argued the math works: “Our AI plus cloud investment has a clear path to attractive ROIC.” Management said AI hardware typically reaches break-even within three years on a five-year useful life, with AI compute supply https://247wallst.com/investing/2026/08/16/spacex-is-planning-to-spend-up-to-500-billion-on-data-centers-heres-why-elon-musk-is-pushing-all-in/ expected to remain constrained industry-wide until at least 2030. That constraint is the whole reason the power, cooling, and networking suppliers behind the buildout keep drawing capital, a group we profiled in a free report on seven AI infrastructure names that aren’t chipmakers https://247wallst.com/pages/ai-power-seven-offer-d905ec99.html . The revenue side supports the case. AI Cloud and Compute Services revenue grew 45%, and AI-related product revenue posted triple-digit growth for the twelfth consecutive quarter. Cloud external growth hit a 22-quarter high, and MaaS annual run rate surpassed RMB 16 billion as of August, tracking a year-end target above RMB 30 billion. What Investors Should Watch Next Shares closed at $130.53 on August 20, up 1.26% on the day and 6.88% over the past week, though still down 10.09% year to date. The setup is straightforward: if Qwen monetization and Zhenwu chip deployments compound as guided, the buyback cut looks like disciplined reinvestment. If AI Labs losses keep widening past RMB 13,861 million without matching cloud margin expansion, the balance sheet, now carrying US$46.5 billion in net cash, becomes the shock absorber. Wu made the trade-off explicit: “It’s only possible to monetize when you have that compute capacity in place.” Contact email protected for any questions or corrections.