Passengers on Lothian Buses have reported a change to automated stop announcements, prompting online speculation that the operator may be using an AI-generated voice.

Commuters have noticed that audio announcements on some services are being delivered in a different tone and cadence, with the voice sounding more natural than previous systems. The change has sparked discussion among passengers, particularly after recordings and comments shared on social media drew attention to the issue.
Passengers on social media compared pronunciations of local place names, with mixed reactions. Some said the announcements were clearer, while others argued the voice struggled with Scots pronunciations, especially on familiar routes.
Examples mentioned included Hamilton Place, which some said was rendered as “ha-meal-ton Place” rather than the locally expected pronunciation.
Another commuter said Hanover Street was announced as “ha-no-ver Street”, with the opening and closing syllables barely audible. Balfour Street was also cited, with one passenger saying it was pronounced as “ball-four Street”, a delivery they described as overly literal.
“It sounds like it was learned from a map rather than from people who live here,” one passenger on social media wrote. Others said the effect was subtle but persistent, with one adding that “once you notice it, you can’t unhear it”.
Several passengers said the change was particularly noticeable because the previous announcements were voiced by Scottish actress Debbie Cannon, best known for her role in BBC Scotland’s River City. Her voice had been used for years and had become a familiar part of daily travel for many commuters.
“The old voice felt local,” one passenger wrote. “You recognised it. This one doesn’t sound like a person who knows the city.”
Some passengers also suggested the difference may be explained by variations within the fleet.
Passengers on social media said older buses appear to continue using pre-recorded human announcements, while newer electric vehicles seem to rely on text-to-speech technology.
While much of the discussion has focused on tone and pronunciation, some commenters linked the change to wider concerns about synthetic voices.
Last year, Midlothian voice-over artist Diane Brooks accused Belgian speech technology company Acapela of using recordings she made more than a decade ago to create a digital voice without her consent.
Brooks said the synthetic voice, known as “Rhona”, was developed using material she recorded 11 years ago, describing the process as “stealing her vocal soul”.
She has alleged that the voice has since been supplied to a number of organisations, including Lothian, though no public confirmation has been provided.
Neither Lothian Buses nor Acapela has commented publicly on the source of the current announcement system, and there is no confirmation that the voice currently in use on Edinburgh buses is connected to the “Rhona” voice.
For many passengers, however, the debate is less about legal or technical definitions and more about how the city sounds. Several said the new announcements were more noticeable precisely because they sit somewhere between human and machine.
“For years you could tune the voice out,” one passenger wrote. “Now it sounds like someone is almost there, but not quite.”
