Amazon Deploys AI System to Gauge Union-Organising Risk in Warehouses
Amazon is using an AI tool called Atlas to assign risk scores that predict the likelihood of union activity at its UK, US and Canadian warehouses.
Materials filed in a legal dispute with Canadian union Unifor reveal that Amazon employs an artificial-intelligence platform named Atlas to evaluate potential labour threats across its warehouses in the United Kingdom, United States and Canada. The system generates a numeric risk score between 100 and 900; a lower figure suggests a greater probability of unionisation, which the company views unfavourably. Atlas categorises sites into five tiers, with tier one signalling imminent disruption and tier five indicating no perceived risk, and the scores inform staffing and resource decisions.
The GMB Union has condemned the tool as "Orwellian" surveillance, while Amazon maintains the AI operates only at the building level and does not track individual employee opinions or affiliations. The document lists two UK facilities—including a Coventry warehouse (BHX4) and a Sutton Coldfield site (EMA4)—as tier-one, alongside fourteen North-American locations flagged as high-risk. Legal experts say the findings contradict Amazon’s public stance that it does not seek to deter union formation.
Why it matters
The story highlights how AI is being used to monitor and potentially suppress labour organising, raising privacy and workers' rights concerns.
In this story
