THE OLD GUARDIAN
Investigative Journalism for the Public Interest
By Christopher Allen
The Old Guardian • 2026
In February 2021, someone decided to trust a document.
The document was called a prospectus. It was produced by TELUS, one of Canada’s largest telecommunications companies, with the help of investment banks on both sides of the border, and it described a business the company was offering to the public for the first time. The business was called TELUS International. It employed fifty thousand people across twenty countries. It had grown revenue at 34 percent per year for three years running. It was, the document said, a “next-generation digitally-led customer experience innovator” — a category-defining company at the intersection of technology and human talent.
The price: US$25 per share. The occasion: the largest technology IPO in the history of the Toronto Stock Exchange.
People read the document. They believed it. They bought in.
Four and a half years later, in September 2025, an independent bank was hired to answer a simple question: what is this company actually worth? The answer came back as a range — between US$3.60 and US$4.70 per share. TELUS paid the shareholders US$4.50 — the top of what the independent expert said was fair — and took the company private.
The people who trusted the document received the most they were legally owed. They still lost 82 cents on every dollar.
Nobody broke any rules. The process worked exactly as it was designed to work. And that, more than anything else, is the TELUS story: a company that followed every procedure, met every obligation, and left the people who believed its pitch holding the bill.
This is the third part of a three-part investigation into Canada’s telecommunications industry. Part One examined Rogers Communications — the outage that took down 911 service for twelve million Canadians, and the regulatory system that looked away. Part Two examined Bell Canada — a company in retreat, cutting its workforce and its ambitions simultaneously. This piece examines TELUS — the smartest of the three, the most sophisticated, and the one whose failures are the hardest to see because they’re wrapped in the most polished language.
I. The Rebrand
In 2023, TELUS International became TELUS Digital. It’s the kind of name change companies make when the old name has stopped telling the story they want told.
“International” said where the work happened. It told you, if you were paying attention, that this was a company built on a global workforce — tens of thousands of people in dozens of countries, doing the unglamorous work that keeps the modern internet running. Moderating content. Labelling data for AI. Answering the phone when something goes wrong.
“Digital” doesn’t say any of that. “Digital” says: think about technology. Think about the future. Don’t think about where the people are.
The company’s own promotional materials describe one of its newer products this way: it “enhances spoken communication in real time by softening accents and improving clarity, without altering the speaker’s voice identity.”
Read that sentence again. We’ll come back to it.
———
II. The Pitch
To understand what happened with TELUS Digital, you need to understand what an IPO is and why companies do them.
IPO stands for Initial Public Offering. It’s the moment a private company decides to sell shares of itself to the general public for the first time — on a stock exchange, where anyone with a brokerage account can buy in. The company gets a large amount of cash upfront. Existing owners can sell some of their own shares at the same time and pocket the proceeds. In exchange, the company takes on a new set of obligations: quarterly reporting, public scrutiny, and the daily judgment of whatever the market thinks the shares are worth.
Companies go public for a few reasons. They need capital to grow. They want to give early investors a way to cash out. Or — and this matters here — they want to lock in a valuation at a moment when the story about the company sounds its most compelling. An IPO is partly a fundraise and partly a pitch. The investment banks running the process spend weeks selling the company’s future to institutional investors before the shares ever trade.
The price the market sets reflects what buyers believe the company is worth — not what it has already built, but what they expect it to build from here. That’s why IPO prices can look dramatic compared to what happens afterward. Investors are paying for the promise.
In February 2021, TELUS took its digital services arm public on the Toronto and New York stock exchanges at US$25 per share. It was the largest tech IPO in TSX history. At that price, the market valued the company at $8.5 billion — more than TELUS itself was worth at the turn of the millennium.
The promise being sold was this: the world’s biggest technology companies need people to moderate their content, train their AI, and answer their customer service calls. TELUS Digital had those people — fifty thousand of them, spread across fifty offices in twenty countries. Google was a client. So was the world’s largest social media company, which TELUS and their lawyers refused to name in the IPO documents even when American securities regulators asked them to. The company had grown revenue at 34% per year for three years running. Profit margins were heading in the right direction. The future was bright.
TELUS raised roughly $490 million in new cash. Its existing shareholders — TELUS itself and Baring Private Equity Asia — sold some of their own shares and collected hundreds of millions more. The public bought in. The deal was done.
Four years later, the same company sold for US$4.50 per share.
———
III. The Flaw in the Foundation
Here’s what the prospectus didn’t say in plain language: the clients driving growth were the same companies building the technology that would eventually eliminate the need for what TELUS Digital was selling.
Google needed humans to label images and train its AI systems. Four days before the IPO closed, TELUS bought a company called Lionbridge AI — a data annotation business that got 65 percent of its revenue from Google. Google also needed humans to moderate content and handle customer service. So did the world’s largest social media company — the one whose name TELUS wouldn’t put in print, even when regulators asked. These were the pillars the growth story rested on.
Both companies were, at the same time, among the most aggressive developers of artificial intelligence anywhere in the world. AI that could label images. AI that could flag harmful content. AI that could handle customer questions without a person on the line.
TELUS Digital’s growth thesis depended on these companies continuing to pay for human labour to do things their own engineers were working, every single quarter, to make a machine do instead. The prospectus listed “client concentration” as a risk, in the careful language that securities lawyers use to describe things without quite saying them. It did not say: our biggest clients are building the technology that replaces us. That sentence does not appear in any prospectus, ever. But it was true, and within two years, it would start to show up in the numbers.
———
IV. The Fall
In the second quarter of 2023, the work started drying up. Not all at once — that’s not how it happens. Large clients began what TELUS would later describe, in the careful language of an annual report, as “cost reduction programs.” Spending commitments were reduced. Service volumes dropped, “more significant than expected, particularly in Europe.” TELUS never named the clients. It never explained why. But the timing tells you most of what you need to know: this is roughly when the major AI labs began deploying the kinds of automated systems that do the work TELUS Digital’s employees had been doing by hand.
By the middle of 2025, the share price had fallen from a 2021 peak of US$35 to US$2.96. That’s a drop of 92 percent from the high, and 88 percent from the IPO price. Adjusted earnings were down 24 percent from the year before. The company wrote off $500 million, because the business it had built — and the price it had paid to build it — turned out not to be worth what TELUS said it was worth when it was selling shares to the public.
In accounting terms, that write-off is called a goodwill impairment. In plain language, it means: we paid too much, and we’re admitting it now, on paper, because the rules require us to.
———
V. The Exit
On September 2, 2025, TELUS announced it would buy back every share of TELUS Digital it didn’t already own.
The opening offer was US$3.40 per share. A committee of independent directors — required by law to act in the interest of the public shareholders, not TELUS — pushed back, and negotiated the price up to US$4.50. An independent bank, hired specifically to answer the question “what is this company actually worth,” came back with a range: between US$3.60 and US$4.70 per share. TELUS paid US$4.50 — close to the top of that range.
Sit with that number for a moment. An independent expert, hired for the sole purpose of being honest about value, said the company was worth somewhere between US$3.60 and US$4.70. Four years earlier, TELUS had sold pieces of that same company to the public at US$25.
The shareholders who bought at the IPO and held on received the most that TELUS was legally required to pay them. They still lost 82 cents of every dollar they put in.
Nobody broke any rules. That’s what makes this worth sitting with. The process worked exactly as it was designed to work, and the result was still an 82 percent loss for the people who believed the original pitch.
———
VI. The Voice
Which brings us back to the accent modifier.
In May 2026, the Globe and Mail reported that TELUS Digital was running its offshore call centre workers’ voices through artificial intelligence in real time — adjusting how they sound before the audio reaches the Canadian customer on the other end of the line. The software comes from a company called Tomato.ai. It takes the worker’s voice, sends it to servers in the cloud, softens the accent, and returns the adjusted audio to the call — all in under a second, without the customer ever knowing it happened.
TELUS’s stated reason is reducing what the company calls “accent-related friction.”
Here’s how one employee described it. According to the president of United Steelworkers Local 1944, a worker based in the Philippines demonstrated the system to a Canadian colleague — turning the accent-softening feature on and off during a call, so the colleague could hear the difference. On. Off. The real voice, then the adjusted one, then the real voice again. The worker reportedly found this funny.
Roch Leblanc, Unifor’s telecommunications director, told Parliament that using AI to deceive Canadian customers about who they’re talking to should be prohibited outright.
The Globe asked Rogers and Bell whether they used similar technology. Both said no, and both said they had no plans to start.
The Globe asked TELUS for comment. TELUS did not respond.
There’s a detail in this story that hasn’t been reported yet, and it matters. Every call that goes through this system is also being routed through Tomato.ai’s servers — which are based in the United States. That means the voice of every Canadian customer on one of these calls, and the voice of the worker they’re talking to, is being processed by an American company’s infrastructure. Whether TELUS discloses this anywhere in its customer privacy terms is not something that has been confirmed publicly. Given how seriously Canadian privacy law treats the cross-border movement of personal data, that’s a question someone with regulatory authority should be asking.
———
VII. The Numbers
Step back now from TELUS Digital specifically, and look at what’s happening to the parent company — because TELUS Digital wasn’t just a business line that went wrong. It was also part of how TELUS told a story about itself that the rest of its numbers don’t quite support.
TELUS employs about 111,500 people. Of those, 25,200 are in Canada. The other 86,300 — 77 percent of the entire workforce — are somewhere else. This is a company that describes itself, in its own annual report, as having contributed $1.85 billion to communities since the year 2000. Three out of every four people who work for TELUS do not work in the country whose name is on the building.
The company’s debt currently sits at 3.4 times its annual earnings before interest and taxes. TELUS’s own target is lower than that — they’re outside the range they’ve told investors they aim to operate within, with no clear timeline for getting back inside it.
In December 2025, the board froze the dividend. Companies don’t usually announce this directly. The language TELUS used was: “There can be no assurance that we will resume dividend increases.” That sentence is doing a lot of work. It is not a pause. It is a company telling careful, sophisticated investors — the only audience that reads sentences like this closely — that the era of automatic annual increases is over, and nobody knows when, or if, it comes back.
Interest costs are climbing, and not because TELUS borrowed more money. It’s because the debt the company took on years ago, at the low interest rates that were common in 2020 and 2021, is now coming due — and has to be replaced with new debt at today’s much higher rates. TELUS expects to pay $1.6 to $1.7 billion in interest in 2026, up from $1.3 billion in 2025. That’s an increase of $300 to $400 million in a single year, and it isn’t going away. It’s the new baseline.
On top of that, TELUS has spent more than $400 million a year on what it calls “restructuring” for three years running, with $500 million budgeted for 2026. At some point, when a company spends this much, every year, on cutting itself down to size, the restructuring isn’t an event anymore. It’s how the company operates.
And the same demographic story that’s hurting Rogers is hurting TELUS too. In its own words, from its own annual report: “decelerated growth observed in immigration has slowed our ability to grow our subscriber base more than anticipated.” Mobile phone net additions — new customers, in plain terms — fell from over 340,000 in 2024 to 207,000 in 2025. The population growth that used to deliver new customers automatically has been switched off by federal policy, and it isn’t coming back on any timeline a telecom company can plan around.
———
VIII. The Comparison
Here’s where this piece earns its place at the end of a three-part series.
Rogers bet on sports, family control, and debt. Of the three companies, it carries the most leverage, has the messiest governance, and produced the most dramatic single failure — the 2022 outage that took down 911 service for twelve million people.
Bell bet on media, and discovered that media was dying faster than anyone at Bell had planned for. Bell is in retreat: selling assets, cutting staff, shrinking toward a simpler version of itself.
TELUS tried something different. Looking at the same landscape — the same slowing immigration, the same dying cable bundle, the same regulatory protections that don’t extend to satellites — TELUS decided the answer was to stop being just a telecom company. Health care. Agriculture. A global digital services workforce. This wasn’t cost-cutting dressed up as strategy. It was an actual attempt to build something new.
Two of those three bets are working. TELUS Health is genuinely growing. TELUS Agriculture occupies a real, if small, niche.
TELUS Digital was the big one — the bet that was supposed to prove the whole pivot made sense. It went from the largest tech IPO in Canadian stock exchange history to a forced buyback at 82 percent below its starting price, in under five years.
All three companies funded their bets with the same source of money: the wireless and cable revenue generated by an industry that, for decades, has faced limited competition and charged Canadians some of the highest prices in the developed world. Rogers used that money to buy sports franchises and take on debt. Bell used it to buy television stations and build fibre networks. TELUS used it to build a workforce of fifty thousand people around the world and sell shares in that workforce to the public.
The protected industry paid for all of it. The protected industry is running out of room.
———
IX. What Broke at the Same Time
This is the thread that ties the whole series together.
The assumptions that Canadian telecom has run on for thirty years didn’t break one at a time, with years in between to adjust. They broke together, in roughly the same two-year window, and none of them are coming back.
Immigration-driven subscriber growth — the engine behind a decade of wireless expansion — was deliberately shut off by the federal government, in response to pressure over housing costs. All three companies say so themselves, in their own financial filings, in plain language.
The cable television bundle — the other pillar the business model relied on — is being abandoned by Canadians at a pace that won’t reverse. Nobody is going back to paying $160 a month for hundreds of channels they don’t watch.
The regulatory protection that’s kept foreign competitors out of Canadian telecom for decades still works on the ground. It does not work in orbit. Starlink’s satellites are already in the sky. The CRTC has no authority over low Earth orbit. Elon Musk doesn’t apply for spectrum licences or negotiate with city councils. Rogers’ own partnership with Starlink — meant to plug coverage gaps — has, in the process, shown Starlink exactly where in Canada the demand for satellite internet exists and what it would cost to serve it directly.
And now, the offshore labour model that TELUS built an entire publicly traded company around is being made obsolete — not by some future technology, but by decisions Google and Meta are making right now, training AI systems to do the work that TELUS Digital’s employees used to do for them.
None of this is reversible on a timeline any of these companies can plan around.
———
X. The Worker
End with the people.
The 86,300 TELUS employees who work outside Canada — in the Philippines, in Central America, in Eastern Europe, in India — some of whose voices are now adjusted by American software before a Canadian customer hears them. Whose identities, on certain calls, have become something a company can switch on and off.
The Canadian TELUS Digital employees who took buyouts in early 2026 — taking with them years of institutional knowledge that, unlike a phone line, can’t simply be rerouted offshore.
The customer on the other end of one of these calls, who has no idea any of this is happening, and who thinks they’re just talking to a person.
TELUS’s stated purpose — printed in its annual report, on its website, in the documents it filed before selling shares to the public — is “to leverage its global-leading technology and compassion to drive social change and enable remarkable human outcomes.”
The brand behind the brands.
The story behind the story.
———
Closing — The Series
Part One ended with a line about walls: the walls only work if everyone agrees they exist.
Three companies. Three investigations. One question running underneath all of it: did Canada’s telecom industry hold up its end of the bargain — serving the public interest in exchange for thirty years of protection from real competition?
The honest answer, after looking closely at all three, isn’t that they failed completely. It’s that they served themselves first, their shareholders second, and the public interest whenever it happened to line up with the first two.
The workers absorbed what was left over. They drove the SIM cards to the towers when the network went down. They took the buyouts when the work moved overseas. They had their voices adjusted by software so a customer wouldn’t notice where the call was coming from.
And the person on the other end of the line was never told any of it.
———
Editor’s Note — November 16, 2026
This article was substantially complete in June 2026. Part Two of The Oligopoly Files published in late spring of that year. What you have just read was the third part, sitting in final preparation.
Five months is a long time to wait. It is also, as it turns out, enough time for the market to deliver its own verdict on the analysis.
On November 9, 2026, Québecor sent an unsolicited offer to TELUS’s board of directors proposing to acquire the company in a cash-and-stock transaction valued at approximately $31 billion. The offer was not new: it was an enhanced version of a private approach Québecor had made in June 2026, the same month this piece was otherwise ready to go. TELUS went public with the offer on November 16, 2026 — the day this article publishes.
The facts, as reported by Global News:
TELUS chair John Manley confirmed the board has received the offer and is conducting a review, but flagged “significant risks,” including the fact that more than four months after its initial approach, Québecor has not explained how the transaction could be approved under the existing telecom regulatory framework. TELUS says it will also consider “strategic alternatives” for shareholders.
The federal government said it has “taken note” of the proposal and that it will be “rigorously reviewed” by the Competition Bureau and the CRTC. Industry Minister Mélanie Joly framed the issue in terms of connectivity, sovereignty, and national security — the same frame Canadian governments have used to justify the oligopoly for thirty years.
The Competition Bureau confirmed it is “aware” of the announcement. Competition lawyer Michael Osborne of Cozen O’Connor identified the central obstacle plainly: Videotron, which Québecor acquired via the Rogers-Shaw restructuring in 2023, holds roughly 12 percent of the wireless market. TELUS holds just under 30 percent. Put them together and you trip the threshold for a deemed substantial lessening of competition under the Competition Act. That is not a technicality. That is a wall.
Québecor, for its part, says the combination will benefit consumers through “more choice, investment and value.” Economists quoted in the initial reporting are skeptical. One pointed to the Rogers-Shaw merger as a reference case. Shaw customers have not seen lower prices.
What to make of all this.
A company under the pressure described in the pages above — carrying debt at 3.4 times earnings against a target of 2.2 to 2.7, issuing junior subordinated notes at 6.75 percent to fund operations, writing down the value of the offshore subsidiary it spent years building, watching customer complaints climb 62 percent while laying off domestic workers and replacing them with software — does not receive an unsolicited $31 billion acquisition offer from a position of strength.
The offer is a signal. Someone looked at TELUS’s balance sheet and its strategic position and concluded the company might be acquirable. That is not a compliment.
Whether the transaction proceeds is a separate question from whether the analysis in this article holds. The regulatory math may well kill it. TELUS’s board may reject it. Québecor may withdraw. The “strategic alternatives” language from TELUS suggests there may be other conversations happening that the public hasn’t seen yet.
The piece you have just read was about a company that made a series of large bets on a vision of itself — the globally competitive technology firm, the customer experience innovator, the humane brand — and found, over time, that the vision did not quite match the numbers. The structure it built, and the debt it accumulated building it, left it with fewer options than it once had.
The November 9 offer did not change that story. It confirmed it.
Iron Ledger classification: PROOF. Source: Global News, November 16, 2026 (Ariel Rabinovitch). TELUS Corporation press release, November 16, 2026. Québecor press release, November 16, 2026. Competition Bureau statement, November 16, 2026. Industry Minister Mélanie Joly statement, November 16, 2026.
———
Sources and Methodology
This investigation draws on the following primary sources:
TELUS Corporation 2025 Annual Report and Management’s Discussion and Analysis, filed with SEDAR+ February 2026. TELUS International (Cda) Inc. Form F-1 Registration Statement, filed with the United States Securities and Exchange Commission, January 2021. TELUS International Initial Public Offering closing press release, February 5, 2021. TELUS Corporation Prospectus Supplement, Series CAR and CAS Junior Subordinated Notes, filed with SEDAR+, April 15, 2025. TELUS International go-private transaction announcement, September 2, 2025. BMO Capital Markets formal valuation of TELUS International (Cda) Inc., delivered to the Special Committee, September 2025, as disclosed in transaction documents filed with the SEC and SEDAR+. iPhone in Canada reporting on TELUS job cuts and USW Local 1944 statements, January 10, 2026. Globe and Mail reporting on TELUS Digital’s use of Tomato.ai accent-modification technology, May 2026. Tomato.ai product documentation, as published on the TELUS Digital website. United Steelworkers Local 1944 press releases and public statements, 2025–2026. Globe Newswire release on Ericsson/Rogers offshoring, October 28, 2025. Statistics Canada and Immigration, Refugees and Citizenship Canada data on population and immigration levels, 2024–2025.
The Old Guardian contacted TELUS Corporation for comment on the financial analysis, the TELUS Digital go-private transaction, the Tomato.ai accent-modification program, and the workforce reduction pattern documented in this article. TELUS did not respond by publication deadline.
The Old Guardian contacted Michael Phillips, President of United Steelworkers Local 1944, to discuss the workforce reduction pattern at TELUS and to request introductions to Canadian TELUS employees affected by the January 2026 voluntary severance program. Phillips did not respond after two weeks.
The Old Guardian contacted Unifor for comment on the outsourcing and offshoring of TELUS telecommunications roles. Unifor did not respond.
Part One of The Oligopoly Files examined Rogers Communications: the 2022 national outage, the classified 911 failure data, and the regulatory system that accepted both. Part Two examined Bell Canada Enterprises: the retreat, the layoffs, and what happens when a company starts selling the things it built.
The Old Guardian accepts confidential tips from TELUS employees, contractors, and former staff at chris@theoldguardian.ca. Secure communications welcomed.


