The Thomsons: How Canada’s Richest Family Built Thomson Reuters and Bet It on AI

Canada’s richest family did not get rich by owning newspapers; it got rich by knowing when to sell them. Over three generations the Thomsons turned a chain of small-town dailies into Thomson Reuters, a company whose products sit on the desks of most large law firms, tax departments and newsrooms in the English-speaking world. The story combines inherited control, professional management and a willingness to exit entire industries when the economics turn. It is part of the Canada Company Stories hub.

Who was Roy Thomson and how did the dynasty begin?

Roy Thomson was a Toronto-born salesman who bought the Timmins Daily Press in northern Ontario in 1934 and spent the next four decades buying newspapers wherever they were cheap. He acquired his first paper outside Canada in 1952, bought The Scotsman in 1953, founded Scottish Television in 1957 and took control of The Times of London in 1966. In 1964 he became Lord Thomson of Fleet, a hereditary peerage that his descendants still hold.

Thomson’s approach was relentlessly commercial. He famously regarded local newspapers as licences to print money because they enjoyed near-monopolies on classified advertising. He did not care much about editorial prestige; he cared about cash flow, and he reinvested that cash in more titles. By the late 1960s he was paying US$72 million for an American chain, Brush-Moore Newspapers, and joining a consortium that found oil in the North Sea.

The North Sea stake mattered more than it looked. It produced the capital that later allowed the family to move out of newsprint and into information services, and it taught the family a lesson it would repeat: own a scarce asset, extract the cash, and be ready to leave when the industry structure changes.

How did Kenneth Thomson turn newspapers into an information company?

Kenneth Thomson, Roy’s son, took over in 1976 when the business was worth about US$500 million, and by the time he died in June 2006 the company was valued at roughly US$29 billion. The transformation came from a deliberate pivot away from print toward specialised information for lawyers, accountants and financial professionals.

The key moves happened in the 1980s and 1990s. In 1978 Thomson bought Wadsworth Publishing, its first step into specialised publishing. In 1989 the various Thomson businesses were merged into the Thomson Corporation. In 1996 the company paid US$3.4 billion for West Publishing, the dominant US legal publisher, a deal that remains the foundation of today’s Legal Professionals segment. Meanwhile the family sold newspapers in waves in 1995, 1996 and 2000, keeping only the Globe and Mail in Toronto as a prestige holding.

Kenneth Thomson was personally reserved and famously frugal, but his management choices were bold. He accepted that the newspapers that had made the family’s fortune were a declining asset and that electronic legal and financial data had better growth, higher margins and far stickier customers. The Thomson Corporation that his sons inherited in 2006 was already an information company with newspapers as a sideline.

Why did Thomson buy Reuters and what came of it?

On 15 May 2007 the Thomson Corporation agreed to combine with Reuters Group in a deal valued at about US$17.2 billion, and Thomson Reuters was created on 17 April 2008. The logic was to pair Thomson’s legal and tax businesses with Reuters’ financial data terminals and global news agency, building a rival to Bloomberg.

The financial data ambition did not work as planned. Bloomberg kept winning terminal share, and in January 2018 Thomson Reuters agreed to sell 55% of its Financial & Risk unit to Blackstone in a transaction that valued the business at about US$20 billion. The unit was renamed Refinitiv, and Thomson Reuters retained 45%. When the London Stock Exchange Group later bought Refinitiv, Thomson Reuters ended up with a large LSEG shareholding that it sold down over several years. In May 2024 the company announced the sale of its remaining indirect stake of about 4.3 million LSEG shares at £91.50 each, for pre-tax proceeds of roughly US$0.5 billion, after which it no longer held any interest in LSEG.

The net result was that the Thomsons used Reuters to enter financial data, decided a decade later that they could not win there, and exited with tens of billions of dollars that funded buybacks, acquisitions and special returns of capital. Reuters News itself stayed inside the group as a smaller, slower-growing but reputationally important segment, generating US$853 million of revenue in 2025.

How does Woodbridge actually control the company?

Woodbridge is a private Ontario company whose shareholders are the descendants of Roy Thomson. Thomson Reuters’ 2025 annual report, filed in March 2026, names Woodbridge as the principal and controlling shareholder, and reporting by The Logic in May 2026 put the Thomson family-linked holdings at about 70.6% of the shares. Because the stake is so large, the family does not need dual-class shares to control the board; a single class of common shares does the job.

David Thomson, Kenneth’s elder son and the third Baron Thomson of Fleet, chairs both Woodbridge and Thomson Reuters. His brother Peter Thomson is also a Woodbridge director. Day-to-day management of the family office passed in September 2024 from long-serving president David Binet to Jay Forbes. Woodbridge also owns 100% of the Globe and Mail, having bought BCE’s remaining 15% in 2015.

The arrangement has consequences for minority shareholders. When Thomson Reuters does a buyback, Woodbridge typically either sells shares proportionately to keep its percentage stable or declines to participate; in the US$600 million buyback announced in February 2026, a spokesperson said Woodbridge would not take part. When the company returns capital through share consolidations, as it did in 2023 and again in May 2026, the family receives the lion’s share of the cash. The May 2026 return of about US$1.44 per share, worth roughly US$605 million in total, would have sent more than US$440 million to family-linked holders if they participated fully.

What did Thomson Reuters report for 2025 and 2026?

Thomson Reuters’ 2025 results, published on 5 February 2026, showed revenue of US$7,476 million, up 3% as reported and 7% organically, with adjusted EBITDA of US$2,936 million and a margin of 39.2%. Adjusted earnings per share were US$3.92, free cash flow was US$1,950 million and net earnings were US$1,502 million. The company raised its annualised dividend by 10% to US$2.62 per share, its 33rd consecutive annual increase and the fifth straight 10% rise.

The “Big 3” segments, Legal Professionals, Corporates and Tax, Audit & Accounting, produced US$6,157 million of revenue with 9% organic growth and a 43.6% adjusted EBITDA margin. Tax, Audit & Accounting was the fastest grower at 11% organic; Legal grew 8% organic and Corporates 9%. Reuters News grew 1% organically to US$853 million and Global Print, the legacy book and looseleaf business, shrank 5% to US$490 million. Some 81% of revenue was recurring and the retention rate was 92%.

Momentum improved through 2026. Second-quarter revenue reported on 5 August 2026 was US$1,954 million, up 9% as reported and 8% organically, with the Big 3 growing 10% organically. Adjusted EPS rose 14% to US$0.99 and free cash flow rose 29% to US$727 million. Management lifted full-year organic growth guidance to about 8% and Big 3 guidance to 9.5% to 10%, while keeping the free cash flow target at about US$2.1 billion. The company also agreed to sell 51% of Global Print to KKR for about US$500 million, a move chief executive Steve Hasker said would let the company concentrate on content-powered AI solutions.

How is Thomson Reuters repositioning around AI?

The company’s answer to generative AI is to own the layer that general-purpose models cannot: verified legal, tax and regulatory content, plus the workflow tools professionals already use. Its flagship is CoCounsel, built on Casetext, which Thomson Reuters bought in 2023 for US$650 million in cash. By 2025 the product family included CoCounsel Legal and CoCounsel for Tax, Audit and Accounting, and the annual report lists broader CoCounsel adoption as the top priority for 2026.

Capital allocation follows that priority. The company says it invested more than US$200 million in AI in 2024 and committed US$150 million to a second Thomson Reuters Ventures fund in February 2025. It bought SafeSend, a US tax-filing software firm, for US$600 million in January 2025, followed by TimeBase, the licensing operations of Imagn Images and Additive AI later in the year, and Noetica in February 2026. In August 2026 Hasker highlighted the release of CoCounsel Legal and what he described as strong evaluation results for the first production-ready version of the Thomson LLM, an in-house model trained on the company’s content.

The strategy is framed around trust. Thomson Reuters calls its approach “fiduciary-grade AI”, meaning tools whose outputs can be relied upon by a lawyer or accountant with professional liability. That positioning is also a defence: if the market decides that any large language model can draft a contract or research case law, a company whose economics depend on per-seat software subscriptions has a problem. The comparable debate around Cohere and Canadian AI sovereignty shows how much of the country’s technology narrative now runs through this question.

Why did the shares fall in 2026 and what did the family do?

Thomson Reuters shares dropped sharply in early 2026 as part of a wider sell-off in software and data companies that investors feared would be disrupted by AI. By early May 2026 the stock was down almost 28% for the year, according to The Logic, and a Motley Fool Canada analysis in September 2026 still described it as down 22% year to date despite a 24% rebound over the preceding three months. The company was trading at about 25 times trailing earnings at that point.

The response was a burst of shareholder-friendly measures. In February 2026 the company announced a US$600 million buyback, which it completed in July. In May 2026 it executed a share consolidation that returned about US$605 million, or US$1.44 per share, to holders. It repaid US$500 million of 3.35% notes in May and agreed the Global Print joint venture with KKR to raise about US$500 million more. Each measure reduced the share count, raised per-share metrics and, because of Woodbridge’s majority stake, moved substantial cash into the family holding company.

Woodbridge’s own behaviour was telling. It sold about US$57.7 million of stock in March 2026 and chose not to participate in the February buyback, which increased its percentage holding slightly. The family’s wealth remains overwhelmingly concentrated in a single listed company whose valuation now depends on how quickly AI tools can be sold into law firms and tax departments.

What role does the Globe and Mail play in the family’s holdings?

The Globe and Mail is the one newspaper the Thomsons kept. Woodbridge owned 85% of the paper for years and bought BCE’s remaining 15% in August 2015, taking full control. Financially it is tiny next to Thomson Reuters, but strategically it anchors the family’s influence in Canadian public life and gives Woodbridge a direct stake in the country’s business journalism.

The Globe has pursued a subscription-first strategy, charging for digital access and investing in data-driven editorial tools, which fits the wider Thomson philosophy of selling information to people who will pay for it rather than depending on advertising. The family has also invested in Canadian media and infrastructure more broadly through Woodbridge, including ownership of a private airport north of Toronto.

For a dynasty whose founder treated newspapers as cash machines, keeping a loss-tolerant national daily is partly a legacy decision and partly a reputational one. It gives David Thomson, who otherwise avoids public attention, a platform of influence without a public profile, and it keeps the family name attached to Canadian journalism even as Reuters remains a global news brand under the corporate umbrella.

How does the Thomson model compare with other Canadian dynasties?

The Thomsons differ from most Canadian business families in two ways. First, they exited their founding industry almost entirely, something the Westons in grocery and the Irvings in oil and forestry never did. Second, they rely on outside professional chief executives, from Richard Harrington in the 1990s to Jim Smith and Steve Hasker today, rather than installing family members as operators.

The family is also unusually concentrated. Where the Desmarais family’s Power Corporation spreads its bets across insurance, asset management and private equity, Woodbridge’s wealth is dominated by one ticker, TRI. That concentration delivered extraordinary returns in the 2010s as the company shed Refinitiv and compounded cash flow, but it also means the family’s fortune rises and falls with investor sentiment on professional software.

Thomson Reuters remains a Canadian-domiciled company listed in Toronto and on Nasdaq, with its operating heart in the United States. That makes it a quiet case study in how Canadian capital can own globally dominant businesses while the operations, customers and most employees sit elsewhere, a pattern also visible in the Brookfield story and in the Toronto-based investors profiled in the City Startup Atlas.

What can founders and investors learn from the Thomsons?

The first lesson is that family control and professional management can coexist. Woodbridge sets capital allocation and chooses the chief executive but does not run product or sales. The result has been three decades of steady reinvention under hired managers, with the family acting as a patient anchor shareholder rather than a dynasty of operators.

The second lesson is about exit discipline. The Thomsons sold newspapers when they were still profitable, sold most of their financial data business when they could not beat Bloomberg, and sold their LSEG stake at a strong price. Each exit was made from a position of strength, before the asset became distressed, and the proceeds were redeployed into higher-margin businesses or returned to shareholders.

The third lesson is that recurring revenue plus proprietary content is a durable combination but not an invulnerable one. Thomson Reuters’ 81% recurring revenue and 92% retention explain why it can raise its dividend every year, yet the 2026 sell-off showed that even the stickiest subscription business is only as valuable as its next renewal cycle. Founders should build the moat and still plan for the day when the moat is questioned.

Finally, minority investors in controlled companies should study how capital is returned. Share consolidations and selective buybacks are efficient for the controlling family; whether they are optimal for everyone else depends on price and timing. Understanding the controlling shareholder’s incentives is as important as understanding the income statement.

Frequently Asked Questions