Are ads the inevitable future of artificial intelligence?

As OpenAI tests advertising integrations for ChatGPT in the United States, the move is being closely watched in Canada, not just by users, but also by marketers, enterprise customers and privacy advocates.

Last month, OpenAI, the company behind ChatGPT, announced it would introduce ads to its free and Go tiers, so "more people can benefit from our tools with fewer usage limits or without having to pay."

In a statement to TechDay Canada, OpenAI said, "We're focused on kicking off our initial testing in the U.S. before any plans to expand."

While there are no publicly confirmed plans to launch this model in the country, it's a topic of conversation north of the border.

The economics are difficult to ignore. Training and operating frontier AI models requires vast computational infrastructure, specialised chips and escalating energy use.

In late 2025, Fortune cited an HSBC Global Investment Research report that projected OpenAI would still not be profitable by 2030.

According to Randall Craig, a Toronto-based digital strategy and marketing expert, the financial reality of AI models makes new revenue streams increasingly likely.

"The cost of AI training is just ginormous," Craig said. "At the very end of the day, the piper has got to be paid."

The trajectory resembles streaming platforms. Early players such as Netflix built their brand on ad-free subscriptions, only to later introduce lower-cost tiers supported by advertising. Craig argues that generative AI platforms may follow a similar path: initial resistance, followed by gradual integration once investor pressure and operational costs converge.

Industry trust under strain

Craig frames the issue in two dimensions: the integrity of AI outputs and the handling of user data.

On the first point, users must believe that model responses are not commercially distorted. If sponsored content appears within or alongside generated answers, even clear labelling may not eliminate suspicion that ranking, emphasis or phrasing could be influenced.

"How do I know that the answers that are provided within the chat are not influenced by the ads that are targeted towards me?" Craig asked.

Even if advertising and model inference operate on technically separate systems, perception matters. For enterprise users relying on AI for research, drafting or analysis, the mere possibility of commercial bias could weaken confidence.

The second issue is data exploitation. Craig connects the debate to what he calls surveillance capitalism - "the idea of another organisation monetizing your identity, your behaviours and your attributes sometimes with and, very often, without your knowledge or without your explicit consent."

Advertising models depend on behavioural profiling. The more granular the targeting, the more valuable the inventory. That dynamic raises questions about how prompts, uploaded documents and usage patterns might be leveraged - even if only in aggregated form.

"People trust ChatGPT for many important and personal tasks, so as we introduce ads, it's crucial we preserve what makes ChatGPT valuable in the first place," said OpenAI in a statement. "That means you need to trust that ChatGPT's responses are driven by what's objectively useful, never by advertising. You need to know that your data and conversations are protected and never sold to advertisers."

Canadian expansion: inevitable or avoidable?

OpenAI has indicated there are currently no plans to expand advertising to Canada. But Craig believes market forces may eventually override geography.

"The answer is, unless there's consumer or business revolt, it's inevitable," he said. A rollout beyond the U.S. would likely depend on early performance metrics, including user retention, subscription churn, advertiser demand, and brand impact. If revenue offsets backlash, expansion becomes more probable. If trust erosion outweighs financial gain, companies may hesitate.

A recent cross-border study from Telus found that 76 per cent of Canadians and 77 per cent of Americans would trust companies more if they reviewed AI systems for potential harms before launching new tools. 

Competitors are taking divergent approaches. Some AI firms have publicly criticised advertising-driven models, positioning themselves as trust-first alternatives. 

During this year's Super Bowl, Anthropic aired an ad spot promoting its ChatGPT rival, Claude. The ad features a man consulting a human-like AI personal trainer for a fitness plan, only for the trainer to pivot into recommending a product complete with a promotional code, a pointed satire of where generative AI could be headed. It closed with a blunt message: "Ads are coming to AI. But not Claude."

Bias, speed and unintended consequences

Craig also warns that AI systems already face scrutiny over training data bias. Introducing advertising into that environment may compound concerns, even if technically unrelated.

For Canadian users, the stakes are familiar. Social media platforms demonstrated how quickly free services can evolve into targeted-advertising ecosystems. The question now is whether conversational AI follows the same arc and whether users will accept that trade-off.

If ads do arrive, the debate will not centre solely on revenue. It will hinge on whether AI systems remain perceived as neutral tools or become another channel in the attention economy.