BCE AI build, fibre gains|Growth or cash-flow squeeze?

Recovery, With a Catch

BCE’s latest quarter looks like an operating recovery, but not yet a clean growth story. Fibre and wireless additions were strong, while the company advanced a large Saskatchewan AI data-centre project that will consume significant capital. The provisional answer is that BCE’s core business is improving, but the stock’s future still depends on whether that improvement can fund a more expensive infrastructure strategy.

Earlier coverage had already changed BCE’s identity for income investors. The company cut its quarterly dividend to $0.44, or $1.75 annually, from $3.99 annually. That reset created room to reduce debt and invest, but it also ended BCE’s old dividend-growth reputation.

The Operating Repair

The latest operating figures support the recovery side of the argument. BCE added nearly 55,000 fibre-to-the-home customers, lifting internet revenue 14.2% year over year. It also added more than 41,000 postpaid wireless subscribers, while postpaid churn fell to 1%, its lowest level in three years. AI-focused businesses grew revenue 29%, although they remain relatively small.

That matters because the dividend cut was supposed to release cash for exactly this kind of repair. BCE paid about $394 million less in dividends during the first half of 2026 than a year earlier and is targeting a free-cash-flow payout ratio of 40% to 55%. For a holder, the trade-off is now clearer: less income growth in exchange for more cash retained inside the business.

The AI Bet

But BCE is not simply using that cash to deleverage. The company has announced $1.3 billion of incremental spending on a 300-megawatt AI data centre in Saskatchewan. The project is expected to be completed in 2027 and, at full run rate, is projected to produce about $500 million in revenue and more than $250 million in free cash flow.

That is the optimistic mechanism. BCE can use its fibre network, enterprise relationships and data infrastructure to move toward higher-value digital services. If the facility reaches the stated economics, it could eventually add a meaningful source of revenue and cash flow beyond traditional telecom.

The near-term mechanism is less comfortable. Construction spending arrives before the facility produces its full earnings contribution. Higher capital expenditures can reduce free cash flow, slow debt repayment and limit future dividend growth. The dividend is safer than it was before the reset, but it is not insulated from the consequences of a poorly timed or poorly utilized expansion.

Signals, Not Proof

There is also a second reading of the quarter. Bell Media benefited from the FIFA World Cup, with operating revenue rising 8.9% to $918 million and advertising revenue increasing 5.3%. Subscriber revenue also grew. Yet BCE’s overall profit was lower year over year. That suggests the headline improvement was not entirely a structural AI or fibre acceleration; part of it came from a major media event that cannot simply be repeated every quarter.

The market’s response remains cautious. One post-earnings analyst compilation raised its revenue outlook modestly but left the consensus price target essentially unchanged at $37.36. The same source projected a sharp decline in statutory earnings per share. That does not prove the strategy is failing, but it shows that better operating metrics have not yet produced a clear long-term re-rating.

The Conditional Recovery

For a holder, BCE now represents a conditional recovery rather than a dependable dividend-growth compounder. For a watcher, the important question is not whether the company can beat one quarter’s expectations. It is whether free cash flow improves after elevated capital spending while debt continues to fall.

The strongest current judgment is that BCE’s telecom engine is showing real traction, while its AI infrastructure strategy remains unproven. The evidence establishes the investment, the expected economics and the recent subscriber gains. It does not establish signed demand, completed construction or realized cash flow from the data centre. The 2027 completion and the project’s actual revenue and free-cash-flow contribution will determine whether BCE has built a new growth engine—or simply exchanged one cash-flow pressure for another.