Consumers Are Learning to Trust AI Through Shopping - PYMNTS.com

The finding comes from the “Agentic AI Report Series,” which tracked 54 personal AI tasks across 5 months from October through February.

In February, 31.4% of AI users turned to generative AI to find product links, the highest adoption rate in the dataset. Month-to-month variance across all five months was just 2.6 percentage points, the tightest of any task measured. Low variance at high volume is the data signature of an established habit, and shopping is where that habit is forming fastest.

The reason product link discovery has taken hold where other tasks haven’t comes down to three factors. It’s high frequency because people shop constantly. It’s low stakes because a wrong result costs nothing. And unlike resume writing or skill building, it carries no demographic preconditions. A 28-year-old comparing headphones and a 64-year-old shopping for a gift are running the same basic task. The PYMNTS Intelligence data confirms it: product link discovery shows broad reach across age groups, income levels and gender, placing it in what the report calls the “truly universal” tier.

That universality is what separates a genuine on ramp from a niche use case. Several of the most discussed AI applications, including resume writing, learning new skills and generating social content, skew toward younger, more educated users. They build deep engagement within a segment but can’t reach the nonuser majority. Product discovery can, and the numbers show it already has.

Key findings from the report include:

• Four percent of AI users used generative AI to find product links in February, the highest adoption rate of all 54 tasks tracked. No other task in the truly universal tier posted both scale and month-over-month growth simultaneously.
• Variance across all five months was just 2.6 percentage points, tighter than any other high-adoption task in the dataset. That consistency separates product discovery from tasks that spike around specific life events or seasonal moments.
• The average active AI consumer now uses 2.69 distinct platforms, with power users averaging nearly four. In a fragmented market, the platform most associated with a user’s first reliable AI task holds a structural advantage.

The competitive picture that emerges from the data is one of fragmentation and opportunity running in parallel. Consumers are spreading their AI activity across multiple platforms, but they’re doing it most reliably around shopping. The platform that owns that experience first has the clearest path to becoming a default.

Shopping was already the most common thing people did online. AI didn’t change that; it just showed up where the habit already lived. Low-stakes tasks are how consumers learn to trust new technology. Shopping is the lowest-stakes AI task in the dataset, and it’s the one with the highest adoption.

At PYMNTS Intelligence, we work with businesses to uncover insights that fuel intelligent, data-driven discussions on changing customer expectations, a more connected economy and the strategic shifts necessary to achieve outcomes. With rigorous research methodologies and unwavering commitment to objective quality, we offer trusted data to grow your business. As our partner, you’ll have access to our diverse team of PhDs, researchers, data analysts, number crunchers, subject matter veterans and editorial experts.