Using AI to mask accents of call centre agents 'misleading,' says union

20,000 telecom jobs lost over past 10 to 15 years, says union detailing concerns before Standing Committee in Ottawa

Using AI to mask accents of call centre agents 'misleading,' says union

Telecommunications unions are urging Ottawa to restrict how artificial intelligence (AI) is used in the industry, saying the technology is reshaping job design, monitoring and performance management.

They warn AI is intensifying worker surveillance, accelerating job losses and enabling practices that could damage trust between employers, employees and customers, according to a report from The Canadian Press (CP).

The Canadian Telecommunications Workers’ Alliance detailed its concerns on April 30 before the House of Commons Standing Committee on Industry and Technology in Ottawa, according to the report. 

The alliance – which includes Unifor, the United Steelworkers and the Canadian Union of Public Employees (CUPE) – represents about 32,000 workers at companies such as Bell, Rogers and TELUS.

Alliance representatives told MPs that AI deployment in telecommunications is already advanced and is affecting both call centre staff and field technicians. They said rapid adoption is raising questions about job security, worker surveillance and transparency for customers.

The alliance is asking for government intervention to set limits on how AI is introduced and used in the sector, including in workforce monitoring and customer interactions.

Accent‑masking and offshoring flagged as risks

In his opening remarks, Roch Leblanc, Unifor’s telecommunications sector director, told MPs he was “aware that at least one company was using AI to mask accents of offshore agents,” according to the CP report, which could "mislead Canadians" into thinking they were speaking with Canada-based employees.

TELUS is the only major carrier reported to be using real‑time accent‑altering AI in its call centres. In its own materials, TELUS Digital frames the technology as “AI‑powered speech enhancement,” saying it partnered with Tomato.ai to “improve clarity and reduce accent‑related friction” while preserving each agent’s “voice identity.”

“Speech-enhancement AI leverages cutting-edge speech-to-speech models to transform audio in real time. These models directly modify the acoustic features of speech, preserving the speaker’s voice while improving clarity and reducing accent-related friction,” the company explains in a June 2025 post on its website.

“The AI works by first encoding the speaker’s voice into a high-dimensional representation that captures both linguistic content and vocal characteristics. The AI modifies only pronunciation-related features before decoding the speech back into audio. This approach allows the solution to address mispronunciations without altering the speaker’s identity or emotional tone.”

Union leaders are concerned that accent‑masking and similar applications of AI could make offshoring less visible while contributing to further reduction of domestic roles in customer service, according to the report.

Previously, an employment lawyer warned that AI in the workplace can be a legal minefield.

Job losses, monitoring and call for federal action

Leblanc argued that customers should be told when AI is involved in handling calls and when work has been offshored. The alliance views this as a matter of both consumer transparency and honest communication about the structure of telecom workforces.

He told the committee that about 20,000 telecommunications jobs have been lost over the past 10 to 15 years due to automation and offshoring. The alliance fears AI will speed up those losses, particularly in routine and high‑volume roles.

He said AI is already being used to track technicians’ movements, measure time spent on tasks and analyse call centre conversations to reroute calls or identify sales opportunities, CP reported. The alliance argues this level of monitoring increases psychological stress and can intensify workloads.

CUPE research advisor Nathalie Blais told MPs the technology can be highly invasive and should be directed to wider public benefit rather than being used in ways that mislead people or cut jobs, according to the report. The alliance is calling for a permanent federal working group on AI, bringing together government, industry and civil society, as well as stronger protections for jobs, workers’ rights and Canadians’ data.

AI Minister Evan Solomon said the federal government’s forthcoming national AI strategy will take labour‑market impacts into account, CP reported.

Canada’s labour protections are not equipped to deal with the speed and scope of job losses linked to AI, leaving workers to absorb most of the costs of technological change, an analysis released earlier this year warns.

In 2023, the federal government released preliminary guidance to federal institutions on their use of generative AI tools for work. The Directive on Automated Decision-Making applies to automated systems, including those that rely on AI, used to influence or make administrative decisions.

Latest stories