The AI Co-Pilot Listening In
For many people, a call to their telecom provider is already a frustrating experience. But for the workers on the other end of the line, there's a new source of stress: an AI system listening to every word they say.
A Telus call centre employee based in British Columbia has spoken out about the company's requirement to use an AI "co-pilot" tool during customer phone calls — part of a broader suite of internal AI products the telecom giant has rolled out in recent years. While Telus frames the tool as a helpful assistant, the worker sees something more troubling: a system being fed real-world data, conversation by conversation, that could one day render her redundant.
"We're essentially training our own replacements," she told CBC News, voicing a fear that is growing across the Canadian call centre industry.
A Familiar Anxiety, Now at the Front Door
The concern isn't new. For years, workers in warehousing, manufacturing, and data entry have watched automation chip away at their roles. But the arrival of large language models and AI voice tools has brought that anxiety directly to service workers — people whose jobs were long considered safe because they require empathy, nuance, and real-time problem-solving.
Call centres employ hundreds of thousands of Canadians, and telecoms like Telus, Bell, and Rogers are among the largest employers in that sector. As these companies invest heavily in AI customer service infrastructure, the workers who have kept those phone lines running are watching closely — and worrying.
Telus, for its part, has not confirmed plans to cut call centre headcount as a result of its AI tools. The company says the co-pilot is designed to support agents, not replace them — surfacing relevant information faster and reducing after-call admin work.
"Support" or Surveillance?
But workers aren't entirely convinced by that framing. The B.C. employee's concern isn't just about job loss — it's also about how the tool changes the experience of work itself. Knowing an AI system is transcribing and analysing every customer interaction adds a layer of monitoring that many find uncomfortable.
Labour experts say this tension is becoming one of the defining workplace issues of the AI era. When a tool is introduced as productivity support but also generates data that could eventually automate the role entirely, the line between assistance and replacement gets very blurry, very fast.
Unions representing telecom workers have flagged AI deployment as a priority bargaining issue, pushing for transparency around how the data collected by these tools is used — and whether it feeds future automation decisions.
What Comes Next
The broader picture across Canada's telecom sector suggests this is likely just the beginning. Companies are under constant pressure to reduce costs, and AI-assisted or fully automated customer service has become a realistic short-to-medium-term option for routine calls — billing questions, plan changes, basic technical troubleshooting.
More complex interactions — complaints, escalations, emotionally charged situations — may remain in human hands for longer. But the volume of calls that actually require that human touch may shrink considerably as AI improves.
For workers like the Telus employee in B.C., the ask feels unfair: do your job well, train the system that could replace you, and hope the company decides to keep you on anyway.
It's a bet many Canadian call centre workers aren't sure they want to make.
Source: CBC News Top Stories
