Who do consumers trust with their data?

This Insights series explores agentic commerce – a transformative shift where AI agents take the lead in shopping, payments and customer decision-making.

AI agents are moving from concept to reality. From booking flights to paying bills and reordering essentials, autonomous systems are being trained to act on behalf of consumers – making choices, completing transactions and even initiating purchases without being prompted.

This is the next evolution of digital commerce: agentic commerce.

To better understand consumer readiness, Worldpay partnered with research firm The Lantern in August 2025. We surveyed more than 2,500 debit and credit card users across the UK, U.S., Australia, Brazil and China. Respondents were introduced to a hypothetical AI agent that could manage shopping, subscriptions and payments – then asked how they felt about using it, which tasks they’d trust it with and who they’d trust to store and manage the sensitive data behind the scenes.

"The winners in agentic commerce will be the ones consumers trust to protect their data."

The results are telling: People are interested, but their willingness hinges on trust. The winners in agentic commerce will be the ones consumers trust to protect their data.

Retailers lead the trust hierarchy

When asked to rate how much they would trust different types of agents to store and use personal data (including payment credentials, ID and address), retailers topped the list in nearly every market.

• In the U.S., consumers gave retailer-provided agents an average trust score of 3.76 out of 5, significantly higher than mobile phone agents like Siri (3.22) and personal AI agents like ChatGPT (2.42).
• In the UK, the pattern held: Retailers scored 3.56, while personal AI lagged at 2.54.
• Even in tech-optimistic Brazil, retailers (3.64) led over mobile (3.29) and personal AI (2.91).

The trust advantage isn’t just about data security – it’s about familiarity. Consumers already transact regularly with retailers and are used to storing sensitive information with them. Extending that relationship to agentic functions feels natural, especially when the brand has a reputation for secure and seamless experiences.

Companies already embedded in the consumer’s day-to-day shopping flow are best positioned to deliver agentic value with confidence.

Younger and male users are more open to agents

Demographics play a role in trust and adoption. According to the data:

• Men were consistently more willing than women to let an AI agent shop for them – especially for high-frequency transactions like grocery, food delivery and entertainment.
• Gen Z (18–24) and Millennials (25–34) expressed the highest willingness to use agents, while consumers over 45 were more hesitant.
• In the U.S. and Brazil, willingness peaked in the 25–34 age group, aligning with digital-native consumer behaviours.

This suggests early adopters will skew younger and more tech-accustomed – a key insight for brands building or piloting AI agent tools. Targeting these demographics could accelerate acceptance and feedback loops.

Top tasks: Bills, transport, subscriptions

When asked which shopping or payment tasks they’d be most willing to delegate to an AI agent, consumers showed a clear preference for low-risk, repetitive transactions.

The top five tasks globally were:

• Paying bills
• Buying public transportation tickets or passes
• Renewing subscriptions
• Ordering food or groceries
• Booking entertainment tickets

More complex or high-stakes purchases – like clothing, electronics or travel – drew greater hesitation. This aligns with broader behavioural insights: Consumers are more willing to trust agents with convenience-driven tasks, but want more involvement in discretionary or high-value spending.

Trust isn’t universal – China stands apart

China was the only market where personal AI agents scored higher than retailers on trust.

• Chinese consumers rated personal AI agents 4.13 out of 5, significantly higher than the global average of 2.63.
• Trust in mobile device agents was also high, at 4.01 – compared to 3.15 globally.

This enthusiasm is likely driven by broader cultural factors, including strong national support for AI innovation and a population that’s already accustomed to digital-first commerce.

However, researchers caution against over-interpreting these numbers. Social desirability bias, national pride and top-down promotion of AI may be influencing responses. Enthusiasm is real – but behavioural follow-through may be more tempered.

Fraud fears are a key adoption barrier

Despite optimism in some segments, most consumers remain cautious. Only 5% globally reported having no concerns about agentic commerce.

"Only 5% globally reported having no concerns about agentic commerce."

Key worries include:

• Fraudulent transactions by rogue agents
• Loss of visibility into where data is stored or shared
• Difficulty reversing unauthorised purchases

These concerns are valid. AI-driven shopping introduces new risks – from account takeover to synthetic identity fraud.

To overcome these concerns, trust must be built into the transaction architecture, not just bolted on afterward.

A new foundation: Verifying identity with Trulioo

To meet this challenge, Worldpay has partnered with digital identity leader Trulioo. As announced in our recent press release, this collaboration embeds digital identity checks directly into the payment flow – giving merchants and platforms a trusted way to validate agents and users alike.

In practice, this means:

• Validating that an AI agent is acting on behalf of a real person
• Verifying payment credentials in real time
• Preventing unauthorised use without introducing friction

This is what trust by design looks like – and it’s essential for scaling agentic commerce safely and globally.

How to build agentic trust

If you're building for this future, three steps matter most:

• Start with trusted brands and channels

  Embed agents into apps or experiences where consumers already have data stored and expectations set.

• Prioritise visibility and controls

  Let users set rules: Approve large purchases, review agent history, pause permissions. Make delegation feel safe.

• Embed proven payment infrastructure

  Use partners like Worldpay with built-in tokenisation, fraud protection and identity tools. Build trust into every layer.

Bottom line: The most trusted win

Agentic commerce is coming fast – but consumers will only follow the agents they trust. That trust will extend to the systems behind the agent: the merchants, platforms and payment providers powering the purchase.

Previously in this series:

Part 4: Mapping the emerging models of agentic commerce

“Agentic commerce” means different things to different people; at Worldpay, we’re helping you cut through that complexity.

Coming up next:

Part 6: What’s stopping agentic commerce?

Despite the hype, agentic commerce faces adoption friction from all sides: consumers, regulators and legacy systems.