An investigation published by openDemocracy claims that Amazon uses an internal data analytics system to assess the likelihood of union organizing at its UK warehouses. The tool reportedly aggregates metrics such as staff turnover, employee complaints, and local demographic data to rank sites by their risk of unionization. Several warehouses, including those in Birmingham, Coventry and Sutton Coldfield, are said to appear in documents reviewed by the reporters.
The practice fits into a broader pattern of algorithmic workplace monitoring already documented in the United States, where Amazon has previously faced accusations of tracking union activity through comparable systems, notably around its Bessemer, Alabama facility. In the UK, the GMB union has run several recognition campaigns at Amazon sites, with mixed results, as workers have at times voted against unionization following intensive internal campaigns from the company.
Amazon has not publicly denied the existence of such tools, though it typically describes its workforce management practices as aimed at anticipating operational needs rather than suppressing organizing efforts. Labor rights advocates argue that using AI systems to predict or preempt collective action raises serious transparency concerns and could run counter to freedom-of-association principles under both UK and international labor law.
The case highlights a growing tension between the use of artificial intelligence in workforce management and workers' collective rights. It also reopens debate over the need for specific regulation of algorithmic surveillance tools in the workplace, an area the UK has so far addressed less strictly than some frameworks proposed within the European Union.