The $80 Billion Takeout War Ceasefire Agreement: Who is Exposed Under the Shadow of AI?

Ceasefire — but only on price

A brutal subsidy war between China's delivery giants has paused, but the real fight is just beginning. It has been reported that Alibaba (阿里), JD.com (京东) and Meituan (美团) — along with challengers like Ele.me (饿了么) and Douyin (抖音) — injected roughly Rmb800–1,000 billion into subsidies over the past year to win share. Markets cheered the pause: Meituan spiked, investors treated a truce as a path back to profitability. Yet the battlefield has shifted from coupons to data, algorithms and last‑mile labor. Who will control the AI entry point — and what does that mean for merchants and riders?

Strategies diverge, costs climb

Meituan (美团) has reportedly decided to cede low‑price orders and defend high‑ticket business (orders >Rmb30), betting that mid‑to‑high‑end users sustain margins. JD (京东) is pushing quality and instant retail. Alibaba (阿里) is pushing Qianwen (千问) as a cross‑service AI entry and has reorganised into an Alibaba Token Hub (ATH) to centralise AI capabilities. Douyin (抖音) is converting content flow into transactions with standalone apps. Labour costs bite: Meituan’s per‑order delivery cost rose to about Rmb7.5 in Q3 2025 and could exceed Rmb9 with full social insurance — automation penetration remains low, so human riders still determine margins.

Data is the prize — and the regulatory risk

The new contest is over who becomes the platform users ask first: an AI super‑entry like Qianwen or the incumbent apps with rich fulfilment networks. Meituan’s advantage is a living map of rider trajectories and billions of order signals; Alibaba’s is an integrated AI front door that can reallocate traffic across services. But this is not only a commercial struggle. Beijing is preparing algorithm rules to prevent anti‑competitive recommendation practices, and data privacy requirements could limit cross‑service stitching of payments and locations. Add US export controls on advanced chips into the mix and the pace of model scale‑up may be constrained by geopolitics as much as by domestic strategy.

Merchants and riders are the hidden casualties

Small restaurants have already borne the subsidy shock — most reported revenue and margin declines during the price war — and now face an AI era where visibility may be decided by opaque algorithms. If platforms prioritise merchants that pay higher fees or better fulfilment economics, will local shops become mere SKUs in someone else’s recommendation list? Riders remain the decisive factor for consumer trust: AI can recommend, but a cold meal at 60 minutes destroys loyalty. The truce on price brought relief. The next round will decide who owns demand, who owns fulfilment, and who pays the ultimate cost. Who can balance automated convenience with human assurance? That will determine the winner.