For years, digital trust largely centered on establishing who was on the other side of an interaction. In an agentic world, identity is still the foundation, but it is no longer enough. We also need to know who or what is acting, whom they represent, what authority they have, and whether the action itself should be allowed.
Consumers already understand this tension. The findings in this report show growing enthusiasm for what AI can do, but clear boundaries around how much control people are willing to give it. They want transparency. They want agency, and as the stakes rise, confidence that AI will act within the authority they intended.
This is where the next chapter of digital trust begins. The question is no longer simply whether we trust AI, but what we are willing to trust it to do.
We cannot blindly trust AI. We must make its actions trustworthy.
Executive Summary
As artificial intelligence takes on greater responsibility in consumers' lives, a new trust question is emerging: how much control are people willing to give it?
Ping Identity commissioned a global survey of 11,000 consumers across 12 countries and found a clear distinction between assistance and delegation. Consumers are increasingly comfortable with AI assisting them, but their willingness to hand over authority to AI varies based on factors such as the task, level of responsibility, and geography.
Greater AI autonomy also raises the bar for trust. Consumers point to the importance of transparency, verified identity, clear authorization, accountability, and human oversight as AI takes on more responsibility.
The results highlight a growing convergence between AI adoption and cybersecurity concerns. Consumers' top worries center on increased cybersecurity risks, personal privacy, a lack of transparency, and identity-based threats such as impersonation, deepfakes, voice cloning, and phishing. Recent industry incidents involving AI agents have demonstrated the real-world consequences of these risks.
In this report, we explore where consumers draw the line when interacting with AI and AI agents, including how their comfort changes as AI moves from providing assistance to taking action. We also examine how the definition of "trustworthy AI" varies by country, providing organizations with insight into how they can strengthen customer trust and engagement.
Darryl Jones, Vice President of Consumer Segment Strategy, Ping Identity
Key Findings: Consumers Are Embracing AI but Expect Trust, Control, and Transparency
Growing Comfort and Trust in AI
Willingness to Delegate to AI
The Need for Human Oversight
Concerns About AI Representation
Where Do Consumers Draw The Line On Autonomous AI Adoption?
AI is evolving from answering questions to taking action, resulting in consumers sending a clear message: they are willing to do more with AI, but granting it control comes with conditions.
For example, the majority of consumers (56%) are most comfortable using AI for support and recommendations, such as troubleshooting a simple task or optimizing a resume. However, comfort narrows as AI moves from providing support and recommendations to taking action, such as preparing a purchase or booking travel. That willingness has clear boundaries, with 25% of consumers indicating they are not comfortable giving AI any level of responsibility.
At the same time, nearly one-third (31%) are already willing to let AI independently make lower-stakes decisions on their behalf, such as scheduling appointments, selecting products, or exploring rentals/homes.
Which of the following best describes the highest level of responsibility you would be comfortable giving AI?
*Percentages do not total 100% due to rounding.
Breaking Down Consumers’ Largest Delegated Authority Concerns
Consumers indicate a distinct line between using AI as an assistant versus delegating responsibility to AI to execute real-world tasks or actions. Many consumers are beginning to experiment with AI agent delegation, but their willingness to hand over control declines as responsibility and risk increase.
The most common concerns consumers cite around autonomous AI are less about technical failure and more about personal agency. They want explicit approval before AI acts, explanations for recommendations, disclosure, human escalation, and assurance that AI represents them accurately.
When considering delegating authority to AI agents, there are four main areas that give consumers pause.
1. Representation
They want to know if the agent will represent them accurately, with 71% of respondents expressing concern about their personality, ethics, or personal values being misrepresented.
2. Authority
Most have a clear spending limit on what they’re willing to let an agent spend without approval, with respondents noting an average of US$200 for routine purchases. However, 42% cap the limit at $100 or less, and 22% refuse to trust AI with any financial transaction.
3. Access
The kind of access they’re willing to grant is also limited, with only 33% indicating willingness to give an agent “read” access to their primary email inbox, even in exchange for saving five hours per week.
4. Oversight
Many are concerned about the level of oversight agents require after task delegation, with 47% more concerned about constantly having to double-check an agent’s work than about the agent making a mistake.
Around the world, attitudes toward autonomous AI delegation vary considerably. Consumers in India (70%) and Indonesia (61%) are substantially more willing to independently delegate decisions and access to AI than consumers in markets like Sweden (21%), the Netherlands (19%), and Australia (18%).
Interestingly, some of the markets most receptive to AI autonomy are also among the most concerned about being misrepresented by it. For example, despite showing high enthusiasm for AI, consumers in India (87%) and Indonesia (86%) are concerned about being inaccurately portrayed by an agent.
How concerned are you that an AI agent acting on your behalf could misrepresent your personality or values to others?
*Percentages do not total 100% due to rounding.
What Do Consumers Expect From Trustworthy Ai?
Security incidents have reinforced the urgency of this shift, focusing the conversation on how AI agents and the identities and permissions behind them should be governed, supervised, and held accountable.
Consumer demand for transparency, verified identity, and data protection has eclipsed the desire for raw speed or model performance. As AI becomes more integrated into daily life, most respondents (56%) stated that knowing when AI is being used is more important than simply receiving the fastest possible result. That expectation can also influence consumer behavior: 20% say they would leave a website immediately if they suspected undisclosed AI, while 20% would avoid using that brand or company in the future.
Consumers also noted they expect the same ethical standards from AI that they expect from human experts, particularly around confidentiality, identity verification, and explainability. They say trustworthy AI should protect personal data, clearly disclose when AI is being used, cite sources, explain recommendations, prove they are interacting with an authentic brand or verified source, and provide a clear path to a human when necessary.
As AI takes on greater responsibility, these findings give organizations a clearer view on what consumers expect from trustworthy AI. This is especially critical given only 12% of consumers fully trust organizations to manage their identity data, down from 17% last year. This declining trust aligns with a growing appetite for oversight, as evidenced by three in four (76%) consumers believing government regulation that protects data in AI applications is important.
The meaning of “trustworthy AI” means different things in different markets. For example, consumers in Australia (68%) and the U.S. (63%) place particularly high value on knowing when AI is being used, while consumers in Indonesia (62%) are much more likely to prioritize speed and accuracy over AI disclosure.
When interacting with a brand (e.g., customer support, personalized shopping), which is most important to you when it comes to AI?
*Percentages do not total 100% due to rounding.
What Organizations Can Learn From Consumers
The findings point to a shift in what digital trust requires. Giving AI greater authority does not eliminate the need for human control; it makes clarity around identity, access and accountability more important.
For global brands looking to build and maintain consumer trust, that means:
As AI gains greater autonomy to act on consumers’ behalf, identity security provides the foundation for greater transparency and control.
The same trust questions consumers are asking today are the ones enterprises must answer as AI agents move from pilots to production. What are you willing to trust AI with?
About the Study
Methodology
This random, double-opt-in survey was commissioned by Ping Identity and conducted by market research company Talker Research, whose team members are members of the Market Research Society (MRS) and the European Society for Opinion and Marketing Research (ESO-MAR). The survey was administered online between June 17 and June 25, 2026, polling 11,000 total consumers across the U.S. (2,000), UK (2,000), France (1,000), Germany (1,000), Singapore (1,000), Australia (1,000), India (500), Indonesia (500), Netherlands (500), Sweden (500), Spain (500), and Italy (500) who have access to the internet.