A new global survey has found that while consumers are becoming increasingly comfortable using artificial intelligence in everyday life, concerns about privacy, transparency and the reliability of AI-generated information continue to influence how much they trust the technology.

Conducted by Talker Research on behalf of Ping Identity, the study surveyed 11,000 people worldwide to examine AI adoption, consumer confidence and attitudes towards allowing AI to make independent decisions.

The findings revealed that respondents knowingly interact with AI an average of 5.6 times per week, although usage varies considerably between countries.

India recorded the highest frequency of AI interactions at 14.2 times per week, followed by Indonesia at 10.5 times.

However, consumers believe they encounter AI more frequently than they realise, estimating an additional 4.3 interactions each week without their knowledge, alongside another 4.3 occasions where they suspect AI is involved.

One of the strongest findings was the importance consumers place on businesses being upfront about their use of AI.

More than half of respondents (56%) said knowing when AI was being used was more important than receiving the fastest or most accurate response. By comparison, just 26% prioritised speed and accuracy over disclosure.

A lack of transparency could also have consequences for brands, with 39% of consumers saying they would become more cautious about sharing personal information if they suspected undisclosed AI use.

Another 20% said they would avoid dealing with the company in future, while the same proportion would immediately leave its website.

Australian consumers were particularly sensitive to the issue, with 24% saying they would abandon a website if they believed AI was being used without disclosure. That figure was slightly higher than the 23% recorded among American respondents.

The research also explored attitudes towards agentic AI, an emerging category of artificial intelligence capable of completing tasks and making decisions with limited human intervention.

Consumers were most comfortable allowing AI to assist with general learning and research (37%), followed by DIY tasks and creative activities such as writing emails or generating images, both at 32%.

Workplace applications were also gaining acceptance, with 24% willing to use AI to summarise meetings or prepare reports. An equal proportion would use the technology to improve resumes or prepare for job interviews.

Other applications included reviewing cover letters (24%), planning meals (23%), researching potential places to live (22%) and providing skincare recommendations (18%).

However, confidence dropped considerably when AI was used for more sensitive matters.

Just 17% were comfortable relying on AI for assistance with minor injuries, while 16% would use it to better understand chronic medical conditions.

That figure fell to 12% for mental health advice and just 10% for prescription medication recommendations.

Consumers were similarly cautious about giving AI control over financial transactions.

On average, respondents said they would hypothetically trust an AI agent to spend $200 on a routine purchase without requiring their final approval.

Despite this, only 31% said they were likely to allow AI to independently make decisions involving activities such as scheduling appointments, selecting products or searching for rental properties.

India again recorded the highest level of confidence, with 70% of respondents indicating they would trust AI to make decisions on their behalf.

The study also highlighted a significant gap between the growing capabilities of agentic AI and consumers’ understanding of the technology.

Only 12% of those surveyed described themselves as very familiar with agentic AI, while 70% were concerned that an AI assistant could misrepresent their personality or values when communicating with others.

Ping Identity Vice President of Consumer Segment Strategy Darryl Jones said greater transparency and user control would be essential as AI agents become more capable.

“People should know when AI is making decisions for them, what they’ve given it permission to do, and when they have the final say,” Jones said.

Despite these reservations, overall confidence in artificial intelligence appears to be improving.

The survey found that 57% of consumers trusted AI to some extent, while 43% said their trust had increased compared with a year earlier.

However, that growing confidence comes with higher expectations around accuracy, accountability and data protection.

Around 40% of respondents reported encountering AI-generated information that was false or fabricated, commonly referred to as hallucinations.

When these mistakes occur, 29% believe the company developing the AI should be held responsible, compared with 15% who blame the person using it.

Consumers were also more likely to consider hallucinations accidental errors rather than deliberate attempts to mislead, with 44% favouring the former explanation compared with 23% for the latter.

The research found that consumers expect AI systems to meet the same ethical standards as human information sources, particularly when it comes to protecting personal data, explaining recommendations and verifying identities.

When using AI for information, 38% preferred responses that presented facts without taking a position, while 21% favoured answers that supported their existing views and 20% appreciated alternative perspectives.

Almost half of respondents (48%) primarily regarded AI as a search engine or information source, highlighting the importance of dependable information as adoption increases.

Jones said consumers should remain cautious about accepting AI-generated responses without verifying their accuracy.

“People should check sources, verify important information and be careful about acting on an AI recommendation without knowing where the information came from,” he said.

The findings suggest that while AI is becoming a routine part of everyday life, businesses developing and deploying the technology will need to prioritise transparency, accuracy and consumer control to maintain public confidence.