Over 1,000 Amazon Workers Push Back on AI

Amazon has been selling AI for the last year as the engine that will drive its growth over the next decade: more intelligent shopping experiences, faster delivery networks, and innovative cloud computing power for governments and large corporations. Within Amazon, however, those involved in developing these technologies say the way they are being rolled out feels less like a careful upgrade and more like a stress test on workers, the power grid, and basic norms.

Late last month, more than 1,000 Amazon employees signed an open letter addressed to CEO Andy Jassy and members of the executive “S-team.” Organized by Amazon Employees for Climate Justice, the letter warns that an “all-costs-justified, warp-speed approach to AI development” could do “staggering damage to democracy, to our jobs, and to Earth.” More than 2,400 workers at Google, Meta, Apple, Microsoft, and other tech firms have also signed in solidarity.

“It’s an aggressive rollout of AI systems,” say the signers of the letter, who range from software engineers and AI specialists to product managers and warehouse workers. They describe a transformation in how decisions are made and how work is measured that is already underway, at the same time as Amazon plans to invest about $150 billion in data center infrastructure over the next 15 years. For them, the core questions are simple: Whose job gets replaced, what powers these models, and where Amazon is willing to draw a line.

The workers are not asking Amazon to abandon AI research and development. Rather, they would like stronger constraints around how it is built and where it is used.

Regarding climate action, the letter calls on Amazon to run all AI infrastructure on clean energy and to stop placing data centers in regions where power grids and water supplies are already under pressure. Workers note that Amazon’s reported emissions have risen about 35% since 2019, despite a pledge to hit net-zero carbon by 2040. In their view, an AI build-out driven by coal and gas would lock in more risk for already stressed communities.

“The letter,” Kevin Roose of The New York Times wrote on Tuesday evening on Twitter, “also asks for new internal guardrails.” Workers demand that there be non-manager forces inside every organization “with actual power to determine if AI will ever be used at their company,” "to decide how any layoffs and hiring freezes related to AI will be carried out,” and “to reduce knock-on effects like carbon emissions and detailed surveillance at the level of individual products.”

AI, layoffs, and tighter metrics

Part of the problem comes with timing. In the past year alone, Amazon has eliminated tens of thousands of corporate jobs while leaders tout AI agents, recommendation engines, and tools like Rufus that could ultimately boost sales by hundreds of millions or even billions of dollars, a trend that employees describe as AI’s justifications for cutting headcount while increasing targets.

In the letter and in interviews with other outlets like Wired and the Guardian, workers discuss internal AI applications which don’t work well yet but are still forced into processes anyway. For workers like those in warehouses, it could mean an increase in their pick rates or more detailed tracking. For technical workers, it could mean using code-writing AI applications and planning assistants while seeing reductions in their teams.

Amazon’s response and why this fight matters

Amazon rejects claims that it has a dangerous strategy when it comes to AI. It cites its commitment to previously mentioned climate goals, its position as the leading corporate buyer of renewable power worldwide, and hundreds of wind and solar power projects to support cloud computing expansion. It’s also been said that AI will lead to new types of jobs and make others more engaging rather than redundant.

In essence, what’s challenged in the open letter concerns if these assurances alone can be sufficient. Since AI systems determine how deliveries occur, how workers get scheduled, and how public services get provided on top of Amazon’s cloud, those building these systems indicate that they need “a voice in these rules.”

For readers though, this article represents the wider context: The dispute at Amazon will not only define how Amazon’s AI strategy evolves. It is an early test of how workers can begin to shift how technologies affect climate policy, labor markets, and democratic processes, before those decisions are codified into actuality.