Why Did Amazon Close Its Shanghai AI Lab After a Billion in Revenue?

Amazon has closed its artificial intelligence research laboratory in Shanghai, ending a six-year run that contributed nearly $1 billion in revenue to the company. Once considered a pillar of Amazon's global AI ambition, the lab’s shutdown underscores shifting priorities amid growing US-China tensions.

The lab, run by Amazon Web Services, was launched in 2018 and was considered one of the company’s most successful R&D hubs outside the United States.

It reached a peak workforce of over 1,000 and developed a neural network framework now used across multiple industries. Yet, in today’s environment, even this level of impact couldn’t shield it from larger geopolitical forces.

Strategic Pressure Overpowered Business Gains

The closure reflects a growing trend of US tech companies rethinking operations in China. Amazon spokesperson Brad Glasser confirmed the shutdown as part of broader layoffs and restructuring within AWS.

He cited the need to make "difficult business decisions" to optimize how resources are deployed and reinvested.

China-based AWS teams were especially vulnerable, given the tightening export laws and increasing suspicion in Washington surrounding cross-border AI development.

Analysts say the decision wasn’t about performance but about risk. The lab's removal demonstrates the deep influence of US policy on global corporate strategy.

Did you know?
Amazon’s Shanghai lab produced over 100 academic papers in six years, many cited globally in graph-based AI research.

Export Controls Hit Innovation at Its Core

Recent US sanctions imposed stricter export rules on the transfer of advanced semiconductors, cloud access, and AI tools to Chinese institutions. These restrictions have hampered collaboration and made foreign AI research centers in China less viable.

Since 2022, researchers in China have struggled to access the specialized chips and software necessary to train cutting-edge deep learning models. Amazon’s retreat is a direct result of this policy shift, signaling that AI innovation is drifting toward geographic silos.

A Broader Pattern of Exit

Amazon is far from alone. IBM closed its Chinese R&D operations in 2024, leading to over 1,000 layoffs. Microsoft wound down its AI and IoT Insider Lab in Shanghai earlier this year. The latter even offered relocation packages to staff willing to move out of China for continued employment.

Other firms, such as McKinsey, have tightened internal restrictions, going so far as to block their China teams from engaging in any generative AI work due to escalating compliance risks. The trend is now clear: strategic exits are becoming routine.

ALSO READ | Burgers, Batteries, and Blockbusters: Tesla’s Diner Has It All

What the Lab Accomplished

Despite political headwinds, Amazon’s Shanghai lab made notable contributions. Lead scientist Wang Minjie described the past six years as “a golden era for foreign R&D in China.”

His team published over 100 academic papers and spearheaded projects in graph machine learning and neural networks for structured data.

One of its most impactful achievements was an open-source graph framework for analyzing relationships in data from e-commerce recommendations to supply chain networks. It reportedly brought in close to $1 billion worth of business applications and enterprise solutions.

Risks Now Outweigh Rewards

Companies once valued talent hubs like Shanghai for their tech depth and cost efficiency. But as Washington increases its focus on national security, economic logic has taken a backseat. US policymakers are wary that breakthroughs developed by local teams could leak into government-backed Chinese agencies.

Amazon’s closure sends a clear message. Even when an operation is thriving, any proximity to sensitive tech sectors inside China brings compliance complexity and future uncertainty. International firms now need to navigate the delicate balance between innovation and regulation.

Global AI Looks Less Connected

The short-term impact is displacement. The displacement leaves researchers, engineers, and support teams unsettled. However, the long-term impact may involve a transformation in the locations and methods of developing artificial intelligence.

Laws or strategic policies now restrict the easy flow of technical knowledge across borders. AI research is no longer neutral. It’s becoming region-locked. Amazon’s Shanghai exit is another indication that the age of globally integrated AI development may be closing.

The next generation of AI systems might develop under separate digital flags. For global research centers, the withdrawal from Shanghai represents a sea change.

Amazon's closure is less about cost and more about caution. It reflects a tech industry rapidly reorganizing itself around the rules of geopolitics rather than the principles of open science.