AI has already overhauled how retailers think about the beginning of consumers' purchasing journey. How AI can help in the middle of that funnel, as consumers deciding what and where to purchase, is starting to take shape.

• Over half of US consumers are comfortable with AI tools carrying out some shopping functions like comparing prices and applying discounts, according to a February survey by Visa and Morning Consult.
• AI agent-led shopping could represent over a quarter of ecommerce spending in the next several years, per a September 2025 Boston Consulting Group (BCG) report.

AI assistance has crept up in usage as consumers discover products and checkout. The Visa study suggests AI is also influencing the middle-game when consumers, having learned about products, determine whether to buy and from which marketplace.

The deciding factor

As shoppers grow more comfortable with using AI tools to search for products, many say they’re also open to assistance in the decision process.

• 58% of US consumers said they’re comfortable with AI comparing prices, per the Visa/Morning Consult survey.

55% are comfortable with AI applying discounts, per the survey.

Retailers can already use AI in the discovery phase through ad placements and app partnerships with chatbots. Utilizing AI at the bottom of the funnel is still being explored. OpenAI pulled back on its Instant Checkout within ChatGPT last month, favoring a framework where the buyer journey is completed on third-party apps.

While platforms continue to define AI's role as a point of purchase, some consumers have grown open to giving agents buying power.

• 38% of consumers said they are comfortable with AI completing a purchase, according to the Visa/Morning Consult survey.
• Still, only 0.3% of US retail ecommerce is expected to come through genAI checkout by 2027, EMARKETER forecasts.

By 2027, AI platform-driven sales in the US are expected to more than double to $44.53 billion, according to EMARKETER’s forecast.

What it means for retailers: Retailers should see that comfort with AI isn’t limited to discovery, many shoppers are now open to AI actively helping compare prices and apply discounts, which are critical decision-stage moments. That creates an opportunity for retailers to integrate AI into mid-funnel experiences where conversion decisions are made, not just top-of-funnel marketing.

Seeking approval

As experimentation has normalized the use of AI discovery shopping tools for many, the majority of consumers still want final approval on purchases.

• 60% would not allow AI to spend any amount without approval, according to Visa and Morning Consult's data.

Just over a quarter (27%) are comfortable allowing AI to spend money autonomously without limits, per the survey.

What this means for retailers: Retailers should design AI-enabled experiences that prioritize transparency and user approval, positioning AI as an assistive layer rather than a fully autonomous buyer. Trust will be a key differentiator: Integrations with established financial institutions and payment networks can help reassure hesitant shoppers and accelerate adoption.

Institutional knowledge

At times, consumers trust old dogs the most with new tricks. Legacy financial institutions instill more trust in shoppers when it comes to AI purchases, particularly if those AI systems are backed by banks themselves.

More than 1 in 3 consumers said they trust bank-backed AI systems (36%) and payment network-enabled AI (35%), per the Visa/Morning Consult survey.

Only 28% said they trust independent AI agents. And 48% of Gen Z respondents said they trust payment network-enabled AI, versus 20% of baby boomers, per the survey.

What it means to retailers: Pushing fully autonomous purchasing too quickly could create friction or erode trust. Partnering with trusted ecosystems (banks, card networks, major platforms) could be a more effective path to adoption than relying on newer, less-established AI intermediaries, especially for older shoppers.

This was originally featured in the Retail Daily newsletter. For more marketing insights, statistics, and trends, subscribe here.