Co-op uses AI to score employee conversations

From quality checks to continuous scoring

According to The Guardian, Co-op’s system records employees’ conversations and assesses their interactions with customers. One worker said monitoring can last for several hours a day, depending on how much time they spend on calls.

The employee described constant scrutiny of every word spoken. They said it felt dystopian, unlike occasional checks for compliance, and feared it could be the start of broader surveillance.

Co-op told The Guardian it rejected the dystopian description, calling AI “an important part of quality assurance processes.” The company said the system is only an aid and does not make decisions.

The practice sits alongside other workplace monitoring efforts:

An estimated 61% of US employers already use AI analytics software to assess employee productivity.

The extra work of being watched

AI changes work not only by automating tasks or replacing employees. At Co-op, it also turns each worker-customer interaction into a data record that managers can evaluate, in the name of optimising processes.

John Chadfield, national representative of the Communication Workers Union, told The Guardian that AI is sold as a solution to every problem. But examples like these, he said, show that it can undermine reasonable working practices and add unnecessary pressure. He argued that unaccountable computer systems should not become people’s managers, and that employees deserve more respect.

I think the key distinction is not whether a system makes the final decision. A tool that scores every conversation can still shape how employees work, even if a person formally remains in charge. What the announcements leave unclear is how workers can challenge those scores or judge whether the monitoring is proportionate. Without that, “support” risks becoming a management system with no clear accountability.