Did Alibaba just kneecap its powerful Qwen AI team? Key figures depart in wake of latest open source release

Source: VentureBeat

Alibaba’s Qwen Team Faces a Leadership Exodus

Alibaba’s Qwen team of AI researchers has long been among the most prolific and well‑regarded in the international machine‑learning community—shipping dozens of powerful generalized and specialized generative models since last summer, most of them entirely open source and free.

But just 24 hours after releasing the open‑source Qwen 3.5 small model series—a launch that earned public praise from Elon Musk for its “impressive intelligence density”—the project’s technical architect and several other team members have exited the company under unclear circumstances. The departures have sparked worldwide questions about the future direction of the Qwen team and its commitment to open source.

Who Left and What Was Said

• Junyang “Justin” Lin – technical lead who grew Qwen from a nascent lab project to a global powerhouse with > 600 million downloads.
• Binyuan Hui – staff research scientist.
• Kaixin Li – intern.

All three announced their departures on X today, without providing reasons or confirming whether the exits were voluntary.

“me stepping down. bye my beloved qwen.” – Junyang Lin (X post)

VentureBeat has reached out to Alibaba for more information and will update when details become available.

The Departing Researchers’ Final Gift: Pocket‑Sized Intelligence

The Qwen 3.5 small model series (0.8 B – 9 B parameters) represents a last masterstroke in “intelligence density” from the founding team.

• Architecture: Gated DeltaNet hybrid, allowing a 9 B‑parameter model to rival the reasoning capabilities of much larger systems.
• Attention Ratio: 3:1 linear‑to‑full attention, delivering a massive 262,000‑token context window while staying efficient enough to run natively on standard laptops, smartphones, and even web browsers.
• Design Philosophy: Lin, a PKU humanities graduate and polyglot, has long championed “algorithm‑hardware co‑design” to bypass compute constraints—a view he detailed at the January 2026 Tsinghua AI Summit.

For developers, Qwen 3.5 wasn’t just another update; it served as a blueprint for the “Agentic Inflection,” where models evolve from chatbots into autonomous “all‑in‑one AI workers” capable of navigating UIs and executing complex code.

The Enterprise Dilemma

“For the 90,000+ enterprises currently deploying Qwen via DingTalk or Alibaba Cloud, the leadership vacuum creates a crisis of confidence.”

• Why Qwen mattered: Many companies adopted it because it offered a “third way”—the performance of a proprietary U.S. model combined with the transparency of open‑weight models.
• Recent restructuring: Alibaba has consolidated its AI efforts into the “Qwen C‑end Business Group,” merging model labs with consumer‑hardware teams. The stated goal is to turn Qwen into the operating system for a new era of AI‑integrated glasses and rings.
• New leadership: The reported appointment of Hao Zhou, a veteran of Google DeepMind’s Gemini team, signals a shift from “research‑first” to “metric‑driven” leadership.

Industry analysts (cited by InfoWorld) warn that as Alibaba pushes to meet investor demands for revenue growth, the “open” in Qwen’s open‑weight models may become secondary—mirroring Meta’s post‑Llama 4 reorganization, which saw the hiring of Scale AI co‑founder Alexandr Wang and the departure of researcher Yann LeCun.

Enterprises relying on the Apache 2.0‑licensed Qwen models now face the possibility that future flagships—such as the rumored Qwen 3.5‑Max—could be locked behind paid, proprietary APIs to drive Cloud DAU (Daily Active User) metrics.

Takeaway: If you value Qwen’s open‑source efforts, download and preserve the models now, while you still can.

The “Gemini‑fication” of Qwen?

The internal friction at Alibaba mirrors tensions seen at OpenAI and Google: the “soul” of the machine often clashes with the “scale” of the business.

“Replace the excellent leader with a non‑core person from Google Gemini, driven by DAU metrics. If you judge foundation‑model teams like consumer apps, don’t be surprised when the innovation curve flattens.” – Xinyu Yang, DeepSeek researcher (X post)

This “Gemini‑fication”—a shift toward a highly regulated, product‑centric culture—threatens the agility that allowed Qwen to surpass Meta’s Llama in derivative model creation.

Junyang Lin served as the primary bridge between China’s deep‑engineering talent and the Western open‑source ecosystem. Without his advocacy, many fear the project could retreat into a “walled‑garden” strategy similar to its Western rivals.

“Leaving Wasn’t Your Choice”

The technical brilliance of the Qwen 3.5 release has been eclipsed by the heartbreak of its creators. Social‑media sentiment among the team reflects mourning rather than celebration:

Chen Cheng, Qwen contributor (X):
“I’m truly heartbroken. I know leaving wasn’t your choice… I honestly can’t imagine Qwen without you.”

Li (X):
“Qwen could have had a Singapore base, all thanks to Junyang. But now that he’s gone, there’s no reason left to stay here.”

What Comes Next for Qwen’s Open‑Source AI Efforts?

• Preserve the existing models while they remain freely available.
• Monitor Alibaba’s roadmap for signals of further commercialization or API‑gatekeeping.
• Watch for community forks or independent stewardship that could keep the open‑source spirit alive.

The departure of key leadership marks a volatile inflection point for Alibaba Cloud and its role as an international open‑source AI leader. How the company balances research integrity with business imperatives will shape the future of Qwen—and, by extension, the broader open‑source AI ecosystem.

The known facts are simple: Qwen has never been technically stronger, yet its founding core has been dismantled. As Alibaba prepares to face investors for its fiscal Q3 earnings report on March 5, the narrative will likely focus on “efficiency” and “commercial scale.”

For the enterprises currently excited about the 60 % cost reductions promised by Qwen 3.5, the immediate future is bright.

But for the larger AI community, the cost of that efficiency may be the loss of the most vibrant open‑source lab in the East.

As Hao Zhou takes the reins, the world is watching to see if Qwen remains a “model for the world” or becomes merely a component in Alibaba’s corporate bottom line.