The growing use of artificial intelligence to monitor employees is raising fresh questions about workplace privacy, accountability and workers’ rights in Britain, after a major UK employer introduced technology that evaluates customer-service calls against more than 50 criteria.
Co-op Legal Services is using an AI model to record and analyse conversations between employees and customers seeking advice about wills, probate and estates, according to a report published by The Guardian on 8 October. The system assigns percentage scores and pass-or-fail assessments that managers can use to examine employee performance.
The development has intensified debate over whether AI-powered monitoring can improve service quality without creating excessive surveillance or undermining trust between employers and staff.
The Co-op says the technology supports quality assurance, coaching and professional development. However, the case raises wider questions about how British employers use automated systems to assess workers, particularly when performance scores may influence managerial judgements.
AI monitoring moves beyond productivity tracking
AI-based workplace monitoring is increasingly capable of analysing more than working hours or output. Depending on how systems are designed, they can assess recorded conversations, identify patterns in employee behaviour and generate performance indicators for managers.
At Co-op Legal Services, the AI system reportedly assesses more than 50 aspects of customer calls. The conversations can involve people dealing with bereavement and other sensitive family circumstances, making the balance between service assessment and employee monitoring particularly significant.
A Co-op spokesperson told The Guardian that the organisation uses AI to review customer conversations and maintain service standards. The company said the technology supports colleagues and helps managers improve the experience provided to clients, while human judgement and accountability remain central to its approach.
The distinction between assistance and surveillance is at the centre of the debate. While employers may use monitoring tools to identify training needs or improve consistency, workers may have concerns about how their conversations are interpreted, how scores are used and whether they can challenge an assessment.
The Co-op case is one example of a broader trend rather than evidence that all British workplaces are using the same technology.
What does UK data-protection law require?
The Information Commissioner’s Office (ICO), Britain’s data-protection regulator, already provides guidance on workplace monitoring.
Its guidance states that employers must monitor workers lawfully and fairly, explain what information they collect and why, and ensure that personal information gathered through monitoring is accurate. Employers should also consider workers’ views or those of their representatives when planning monitoring systems, unless there is a good reason not to do so.
The regulator warns that excessive monitoring can interfere with workers’ privacy and negatively affect their mental wellbeing. It also highlights the importance of challenging inaccurate information when monitoring data is used in performance reviews.
These requirements are particularly relevant when AI-generated assessments influence decisions about employees. A score may appear objective, but its usefulness depends on the quality of the underlying data, the criteria selected and the context in which a conversation takes place.
The ICO’s guidance on automated monitoring also addresses human oversight and the safeguards required when organisations use automated processes to manage performance.
However, the existence of regulatory guidance does not establish whether a particular employer has breached the law. No such conclusion should be drawn about Co-op Legal Services without evidence from the regulator or another competent authority.
Government faces questions over workplace safeguards
The debate has also reached Westminster.
In a written parliamentary answer published on 8 September 2026, Baroness Lloyd of Effra said the government was consulting on workplace monitoring technologies to help ensure that their use remained fair and transparent while workers’ rights and protections kept pace with technological change.
The consultation closed on 30 September 2026. Its outcome could help shape the future policy framework for monitoring technologies in British workplaces.
The government has also highlighted the need to build an evidence base on how AI affects employment, skills and productivity. Yet measuring productivity alone may not capture the wider consequences of automated assessments, including employee confidence, privacy and the quality of workplace relationships.
A key question for policymakers is whether existing safeguards are sufficient when monitoring becomes continuous and assessments are generated by systems that employees may not fully understand.
Can AI assess empathy?
The issue becomes more complicated when employers use AI to evaluate communication skills rather than simply count completed tasks.
In legal services, customer conversations can involve grief, uncertainty and emotionally difficult decisions. An automated system may identify whether certain phrases were used or whether a required procedure was followed, but those indicators do not necessarily capture the full context of a conversation.
For example, a customer dealing with a recent bereavement may respond differently from someone making a routine enquiry. A rigid scoring system could risk overlooking contextual differences unless it is carefully designed and reviewed.
This does not mean AI cannot contribute to quality assurance. It means that organisations need to explain how their systems work, test whether assessments are reliable and ensure that managers do not treat automated scores as unquestionable evidence of performance.
Workers should also have meaningful opportunities to question assessments they believe are inaccurate or misleading.
What comes next?
The central issue is not simply whether employers should use AI to monitor employees. It is whether such systems operate with sufficient transparency, proportionality and human accountability.
For British workers, important questions remain: Are they clearly informed about what is being recorded? Can they access and challenge their assessments? How long is the information retained, and who can use it? Are employers evaluating the effects of monitoring on staff wellbeing as well as business performance?
For regulators and policymakers, the challenge is to balance legitimate business needs with the rights of people whose work is increasingly assessed by automated systems.
As AI becomes more embedded in everyday management, the standards governing its use could help determine whether the technology becomes a useful tool for professional development or a source of persistent workplace pressure.
The Co-op case provides a timely starting point for that debate. The wider question for Britain is whether employees can benefit from AI-driven workplaces while retaining meaningful privacy, dignity and a fair opportunity to challenge decisions about their performance.
