The quiet expansion of AI-powered workplace surveillance is no longer a future-of-work hypothetical — it is a present-tense reality affecting workers across sectors. Two recent UK cases illustrate the pattern: tech training company Multiverse deployed an AI monitoring system that instructors described as "remorseless" and "unnerving," while The Co-Op introduced automated listening technology that rates every customer phone call made by certain employees. The through-line in both cases is asymmetry. Management selects the system, defines the metrics, and owns the data. Workers discover the monitoring after implementation, often with no meaningful input into what counts as a good phone call or a productive training session. The AI becomes a black-box supervisor whose judgments are difficult to contest. Multiverse instructors went public — a genuine whistleblowing act in a sector where employers hold significant leverage over staff. That they felt compelled to use the language of alarm ("remorseless") rather than mere inconvenience signals that the experience of being continuously scored by an opaque system is qualitatively different from traditional performance reviews. The Co-Op case is structurally more common and arguably more concerning. Automated call-scoring systems are cheap, scalable, and already widespread in customer-service operations. They promise efficiency gains for management but impose a constant performance tax on workers who know every interaction is being graded by a system they cannot see, question, or appeal. The regulatory picture remains thin. The UK's data-protection framework nominally covers automated decision-making, but enforcement lags far behind deployment. The EU's AI Act introduces risk categories for workplace AI, but implementation timelines stretch into 2026 and beyond. In the meantime, the surveillance stack grows. What makes this extractive rather than generative is the unidirectional flow of information and power. If AI monitoring genuinely improved working conditions — flagging burnout, reducing abusive customer interactions, distributing workloads more fairly — workers might welcome it. Instead, the pattern so far is one-way accountability: workers are watched, scored, and ranked while the systems themselves face no equivalent scrutiny. The Guardian's open call for worker testimonies is itself a signal. When a major newspaper asks readers to self-report surveillance experiences, the implication is that official channels — unions, regulators, HR departments — are not surfacing the problem at the speed it deserves.
