Alibaba's Partnership Committee Rebukes DingTalk Management After a Viral Resignation Essay

Alibaba's Partnership Committee, the highest-level governance body inside the company, circulated an internal memo on June 10 directly criticizing the management practices at DingTalk, its enterprise collaboration subsidiary. The trigger was a resignation essay from a former DingTalk product manager that spread quickly across Chinese social media. For a group-level body to comment on the internal culture of a single business unit is unusual, and the specifics of the case say more about the realities of building AI products under deadline pressure than the corporate-values language of the memo lets on.

What happened

The essay, titled "Inside DingTalk," was written by a product manager who joined the company in June 2025 and worked on DingTalk's flagship AI product, internally called "ONE" (Project O). ONE shipped as part of DingTalk 8.0 and was framed as the platform's central AI effort: an AI-powered work information stream meant to surface relevant tasks, messages, and documents instead of leaving users to navigate them manually. The author described how, after launch, the product ran into shifting internal priorities and escalating pressure from leadership.

The Partnership Committee's memo, titled "Loyalty and Growth: That's Alibaba Culture," said the practices described in the essay "have never been the direction Alibaba's culture advocates," and asserted that "mutual respect, treating people as people, loyalty and righteousness" form the basis of the company's culture. It went further on management philosophy, stating that innovation in the AI era depends on "employee passion and creativity," not "pressure and mechanical execution," and that managers are responsible for articulating vision, owning accountability, and motivating their teams.

The controversy landed against a backdrop of leadership churn. Chen Hang, who goes by the alias Wuzhao and founded DingTalk, returned as CEO in April 2025, replacing Ye Jun. Former DingTalk Vice President Ma Ruila published his own essay, "Outside DingTalk," reflecting on his departure. You can read more about DingTalk and its product line at dingtalk.com.

What's actually new here

Resignation essays from Chinese tech workers are not new, and neither is internal turmoil at large platforms. What stands out is the level at which Alibaba chose to respond. The Partnership Committee is the body that historically controlled board nominations and embodies the company's governance structure. It typically does not weigh in on the day-to-day management of individual products. Doing so publicly, even via an internal memo that predictably leaked, signals that the group treated this as a reputational and structural problem rather than a routine personnel matter.

The other notable element is the subject. ONE is positioned as DingTalk's core AI bet, the product meant to justify the platform's pivot toward AI-driven enterprise services. The essay ties the management dysfunction directly to the pressure of delivering on that bet. That connection is the part worth paying attention to, because it reflects a pattern visible well beyond Alibaba.

The pattern behind the memo

Enterprise software companies across the industry have spent the last two years reorganizing around AI features, often with timelines set by competitive optics rather than by what the underlying models can reliably do. An "AI-powered work information stream" is a hard product. It requires ranking and summarization that stay useful across messy real-world enterprise data, latency that does not annoy users, and accuracy that does not erode trust the first time the feed surfaces the wrong thing. When leadership sets aggressive launch dates and then shifts priorities afterward, as the essay describes, the people closest to the product absorb the gap between what was promised and what the technology delivers.

The memo's framing, that AI-era innovation comes from passion and creativity rather than pressure, reads as a corrective. Whether it changes anything is a separate question. Cultural memos are cheap; the incentives that produce the described behavior, namely the pressure to demonstrate AI traction to executives, investors, and the market, are not addressed by a statement of values. The gap between the two is where this story actually lives.

What changes

In the short term, probably little at the product level. ONE remains part of DingTalk's roadmap, and the company's AI strategy does not hinge on one product manager's departure. The more meaningful effect is on how DingTalk's leadership operates under closer scrutiny from the parent group. Chen Hang returned to run the subsidiary roughly a year before this episode, and a public rebuke of the unit's management culture puts that leadership transition under a brighter light.

For anyone tracking the enterprise AI market in China, the episode is a reminder that the constraints on these products are organizational as much as technical. The benchmark scores and feature demos tell one story; the internal essays tell another, and the second one is usually closer to how the software actually gets built. Alibaba's response acknowledges as much, even if the memo stops short of naming the incentives that created the problem.