Amazon uses an in-house AI forecasting tool called Atlas to score its warehouses on the likelihood that workers will unionise, stage sit-ins or carry out other "disruptive labour events", according to internal documents reported by openDemocracy and circulated widely on Hacker News this weekend.
The documents, disclosed as part of a legal battle between Amazon and the Canadian union Unifor, suggest the retail giant amasses "hundreds of data points" — including posts on Reddit forums, employees' responses to internal feedback surveys and records of previous industrial action — to award each warehouse a union-activity risk score between 100 and 900. The lower the score, the more likely Atlas believes workers are to organise, openDemocracy reported. For Amazon's leadership, the document states, higher scores are "better".
How the Atlas Risk System Works
According to the appendix describing the tool, warehouses are sorted into five tiers based on their Atlas score. Tier 5 means "no immediate perceived risk", while Tier 1 means "future activity is imminent". The document says the scale can be used to "inform staffing decisions, resource planning, standard actions, and ownership to mitigate risks to Amazon labour structure".
Sixteen Amazon warehouses across the US, Canada and the UK are ranked Tier 1, according to the document. Two of them are in Britain: a £500 million robotic fulfilment centre in Sutton Coldfield, north-east Birmingham, known internally as EMA4, and the warehouse near Coventry where 62-year-old Stuart Hylton works.
Hylton, a GMB union member who receives and dispatches goods at the Coventry site, told openDemocracy he learned about Atlas from the union and believes the depth of data collection creates a severe power imbalance. "It's not right in a democracy… the employer should not be interfering," he said. "Workers should be allowed to decide if they wish to choose to unionise."
Amazon: Building-Level Data, Not Individual Surveillance
An Amazon spokesperson refuted the characterisation of Atlas as "individual employee monitoring or surveillance", stating: "This is site/building level activity only." The company said it does not track individual employees' opinions, affiliations or union membership, and does not use Atlas to monitor dissent but to "help us understand where associates need more support and to ensure we're addressing issues that they raise".
Amazon also told openDemocracy that a framework called Atlas is not currently used in the UK, though the company had not answered at publication whether a similar system runs under a different name there.
For ongoing coverage of workplace AI and the AI industry news that shapes it, AI Buzz Wire tracks every major development.
Documents Surfaced Through a Canadian Union's Legal Fight
The Atlas materials are training resources for Amazon managers in Canada and were disclosed during litigation brought by Unifor, which was later shared with the US non-profit publication Truthout. The British Columbia Labour Board ultimately ruled that Amazon had violated multiple local labour laws, following an earlier finding that the company had unlawfully imposed a wage freeze on unionised workers at an outsourced warehouse.
Michael Newman, a partner at London law firm Leigh Day, told openDemocracy the documents tell "a very different story" from Amazon's previous public assertions that it does not seek to prevent workers from joining a union. "Now we know Atlas is used by Amazon, we'll be asking more questions about what data points are used, how these scores were reached," he said.
Leigh Day and the tech justice non-profit Foxglove are already representing two workers in a legal challenge against Amazon's "click-to-quit" QR codes, which they argue were used to pressure staff into ending union membership. Foxglove's co-executive director Rosa Curling described the surveillance detailed in the Atlas document as "highly intrusive", saying it shows a "very intentional policy to try and bust organising" — and that her organisation is reviewing the documents to see whether they strengthen its existing claims.
Coventry: The UK's Test Case for Amazon Organising
The two British Tier 1 sites sit at the centre of the UK's most serious push to unionise Amazon. Workers at the Coventry warehouse became the first in the country to take strike action against the company in January 2023. Six months later, around 2,600 employees voted in a formal ballot for union recognition, which GMB lost by just 28 votes — a margin the union and Foxglove have blamed on what they call "dirty tricks", including mass hiring of non-unionised workers and QR codes displayed on screens, cafeteria tables and toilet doors in the run-up to the vote.
The Atlas revelations land in a familiar regulatory context. In 2022, France's data watchdog CNIL fined Amazon €32 million (£27 million) for breaching GDPR by tracking warehouse staff so precisely that alerts triggered when workers were inactive for as little as one to ten minutes. Asked about the AI tool, a UK government spokesperson said employers and workers share an interest in workplace monitoring being used "transparently and in ways that are fair, accountable and reliable", adding: "We expect all employers to respect workers' rights to take part in lawful trade union activity."
Whether regulators treat warehouse-level AI risk scoring as routine business analytics or as an interference with the right to organise is now the question at the heart of two live legal challenges — and, with the documents public, Amazon's answer is being tested in court rather than in a press release.
---
Stay Informed on AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
