Marketers Bet on AI Assistants, but Only 23% of Consumers Trust Them

Marketers are moving budget toward the place they expect customers to go next: the AI assistant that answers a question about which product to buy. The assumption underneath is that shoppers will treat those answers the way they once treated a search result. The newest large survey suggests they do not.

The gap between where the money is going and where confidence sits is wide, and a second study shows it is already costing brands customers. Together they make a case that trust, not reach, is the scarce resource in AI-driven discovery.

Budgets Are Moving Faster Than Belief

Kantar's Media Reactions 2026, published on September 29 and based on 806 senior marketers and 23,452 consumers across 33 markets, reports that a net 75% of marketers plan to raise investment in AI assistants in 2027. Only 23% of consumers say they trust recommendations from AI assistants, and just 32% use them to research brands or products at all.

Kantar frames this as a sign of unfinished work rather than a verdict on the channel. Gonca Bubani, the firm's Global Director, Media, put the risk plainly: "There's a lot of money moving towards AI on the assumption people will use it the way the industry expects." Meanwhile 62% of marketers expect generative AI to play a pivotal role in brand recommendations, so the expectation is running well ahead of observed behavior.

The same study shows marketers are less sure of their footing generally. Only 51% feel confident extracting actionable intelligence from their data, and confidence that the organization has the right media mix fell to 57% from 64% a year earlier. At the same time, 66% plan to invest in micro-creators, a sign that human voices are part of the hedge.

Distrust Already Has a Price Tag

The Usercentrics State of Digital Trust in 2026 report, built on 11,000 consumers across seven markets surveyed in March 2026, puts numbers on what mistrust does. 47% took at least one action with direct revenue consequences in the past six months because of how brands handle their data in AI: 24% canceled a subscription, 20% switched to a competitor and 20% reduced their spending.

Personalization is where the friction shows. 71% find AI-driven personalization intrusive, and 52% trust AI less than humans with their personal data, up from 48% in 2025. Behavior at the consent banner is shifting too, with 48% saying they click "accept all" less often than before. For marketers, that means the data feeding an assistant or a personalization engine is getting harder to collect and easier to withdraw.

Transparency Is the Part That Pays Back

The Usercentrics data also shows an upside. 52% of consumers would pay more for brands they trust on AI, at an average premium of 7%, and 53% of privacy-aware consumers are comfortable with personalization compared with 19% of those who are not. The three things consumers ask for have not changed in two years: clear explanations of data use (44%), strong security guarantees (42%) and control over what they share (41%).

That is an operating brief rather than a legal one. The brands most likely to be recommended and believed are those that can explain, in plain language, what an AI system knows and why it said what it said.

A Playbook for Earning Trust Before You Scale Spend

• Test before you commit budget. Run small, measured pilots on AI assistant visibility and track whether recommendations lead to real conversions, as Kantar advises, before shifting a large share of spend.
• Map the pre-conversion journey. With only 32% of consumers researching through assistants, find where the rest still form opinions and make sure your brand shows up there.
• Explain your data use in plain language. Put a short, readable account of what you collect and why next to every personalized experience, since 44% want exactly that.
• Give people control. Offer preference settings and easy opt-outs, and treat consent rates as a trust metric rather than a compliance chore.
• Pair AI output with human proof. Use creator voices, real footage and named experts to back up claims a machine could have made up.
• Audit your trackers. Know what each pixel and script collects and where it goes, because the same consumers who distrust AI are watching how their data moves.

Marketers do not have to choose between AI and trust. But the evidence from 23,000 consumers says the order matters: earn the belief first, and the spend that follows has something to stand on.