NMI Research Reveals 70% of Consumers Demand Final Say Over AI-Driven Purchases

New research into consumer sentiment reveals a stark disconnect between the industry’s push for autonomous AI agents and the actual comfort levels of American shoppers. For fintech professionals, this gap highlights a critical friction point: while AI is successfully streamlining discovery and search, the final payment remains a strictly human-guarded gate.

What was announced

NMI, a global provider of embedded payments infrastructure, released its Embedded Payments: Agentic Commerce Reality Check report, based on a survey of 1,000 U.S. adults. The findings suggest that while the industry is pivoting toward "agentic commerce"—where AI agents browse and buy on behalf of consumers—shoppers are currently drawing a hard line at the point of transaction.

The data shows that 53% of U.S. consumers have used AI for research or information searches in the past month, and 33% have used it for product discovery. However, only 11% have used AI to complete a transaction. Trust remains the primary barrier; 69% of respondents do not trust AI to process payments securely, and 70% insist on the ability to review or override an AI’s decision before any money changes hands. Only 10% of consumers are currently willing to grant AI full control over a purchase.

Demographic trends show faster adoption among younger cohorts and parents. Nearly half of parents (48%) have used AI for shopping discovery, compared to 25% of adults without children. Furthermore, 48% of both Millennials and Gen Z expressed excitement about agentic commerce, yet even among these groups, only 14% and 13% respectively would grant an AI agent full purchasing authority. The survey also noted that 55% of consumers are uncomfortable with AI recommending which payment method they should use.

"Payments is where the promise of agentic commerce meets real consumer risk. If an AI agent is going to spend someone’s money, it has to be clear what that person agreed to, what the agent is allowed to do and when it needs to ask for approval again. That means payment infrastructure will need to carry much more context. It will have to connect identity, authentication, permissions and secure credentials so that every transaction reflects what the consumer actually intended, not simply what the AI was technically able to do."

Tiffany Johnson, Chief Product Officer at NMI.

The companies involved

NMI is a global leader in embedded payments infrastructure, providing the underlying technology that enables software vendors and ISOs to integrate payment processing into their own platforms. The company focuses on the full commerce lifecycle, from sign-up to payout, across physical, mobile, and online environments. By positioning itself at the infrastructure level, NMI facilitates the flow of transaction data and security credentials that are essential for modern retail environments.

Salesforce, referenced in the research regarding the 200% rise in AI-driven search, is a dominant force in the global CRM and enterprise software market. The company has increasingly integrated artificial intelligence across its suite of products to automate customer service, sales, and marketing workflows. As a major player in the digital transformation space, Salesforce’s data on consumer behavior often serves as a benchmark for how quickly AI-driven tools are being adopted at the top of the sales funnel.

What FF News has reported before

FF News has tracked the steady integration of AI across the financial services and insurance sectors. Recent coverage includes the appointment of a former Salesforce AI leader to drive autonomous revenue cycles in Waystar Appoints Former Salesforce AI Leader Amit Khanna as CPTO to Drive Autonomous Revenue Cycle. We have also reported on the acceleration of AI in the underwriting space through partnerships such as Sixfold and Sollers Consulting Partner to Accelerate AI Underwriting Transformation for Global Insurers and the launch of the Appian and Synechron Launch AI-Powered Open Underwriting Stack for Insurance Carriers. Additionally, the industry's focus on these technologies was highlighted in our coverage of BrokerTech Connect Chicago 2026: BTV Unveils Insurtech Conference Lineup and AI-Focused Sponsors.

What this means

The industry is currently facing a "trust tax" on autonomous commerce. While software providers are racing to build agents that can execute the entire shopping journey, the market is signaling that it only wants a co-pilot, not an autopilot. This puts significant pressure on payment gateways and infrastructure providers to develop more granular permissioning systems. The move to agentic commerce will likely stall unless the sector can move beyond simple "buy" buttons to sophisticated "conditional authority" frameworks. The challenge for the next year will not be the technical capability of AI to find a product, but the ability of the payments stack to prove to a skeptical consumer that their wallet is safe under AI management.