Telus Uses AI to Alter Call-Agent Accents in Real Time
Reports from The Globe and Mail and iPhone in Canada indicate that Telus, via its Telus Digital division, is implementing AI to alter the accents of call-centre agents in real time. This speech-to-speech technology, developed by Tomato.ai, aims to minimize what Telus refers to as "accent-related friction" among offshore agents. Labour organizations have condemned this tactic as misleading, calling for mandatory transparency. In contrast, Rogers and Bell have stated they do not intend to use similar technology. The introduction of this system has sparked significant public outcry in Canada. The technology employs automatic speech recognition, accent conversion models, and neural vocoders, raising concerns about privacy and worker rights in customer service contexts.
Key facts
- Telus is using AI to alter call-agent accents in real time via its Telus Digital unit.
- The speech-to-speech tool is built by Tomato.ai.
- The technology is applied to offshore agents' voices to reduce 'accent-related friction'.
- Labour groups have criticized the practice as deceptive and urged mandatory disclosure.
- Rogers and Bell told The Globe and Mail they have no plans to adopt similar technology.
- The rollout has provoked swift public backlash in Canada.
- Voice conversion at low latency combines ASR, speaker/accent conversion models, and neural vocoders.
- Real-time speech-to-speech systems can mask speaker identity and smooth prosodic or phonetic patterns.
Entities
Institutions
- Telus
- Telus Digital
- Tomato.ai
- Rogers
- Bell
- iPhone in Canada
- The Globe and Mail
Locations
- Canada