Earning Trust in AI

Even while artificial intelligence becomes a greater part of everyday life, consumers remain skeptical of the fast-developing technology. A 2026 Pew study found that more than half of Americans expressed a lack of faith in U.S. companies to develop and use AI tools responsibly. 

As AI capabilities converge, the enterprises that win will be the ones that secure something far more elusive than new business: trust. That’s exactly what kyu companies are doing.

Here are four ways they’re gaining confidence with users while making the most of new technology.

Testing AI the way people will actually use it

IDEO has long used real, personalized data, rather than hypotheticals, in its research process. Now, it uses generative tools to make that approach easier to scale. In one recent project for a financial services client, the team used AI to generate personalized prototypes that pulled data and context from each participant’s interview, which drew out more honest feedback.

That kind of feedback is essential to understanding whether a new product will properly serve its end user. ​“What we’re designing won’t just succeed or fail based on responsiveness, server uptime or the quality of the information hierarchy design, but both how much and how the system understands and can make sense of you and your context,” says Jenna Fizel, Managing Director, Global Head of Emerging Technology at IDEO.

Seeing how people respond to an AI system is also how designers learn what earns their trust. In the U.S., Fizel says, ​“there is huge skepticism, distrust, and anger as the default stance for many people,” which can make proof points and caution critical for building AI tools. As people anthropomorphize AI, a system’s conversational tone matters too, because it can create ​“expectations of functionality, capability and power you may or may not mean to imply,” Fizel adds. ​“This is one of the reasons why a lot of customer service chatbots are still infuriating.”

IDEO has formalized this thinking in a human-centered framework for using AI in design research. It treats AI’s speed and scale as dials to adjust, so teams can add friction when a project calls for it, keeping people in charge of the most important decisions.

Showing the human impact of new technology

“Many of our clients are technology companies working with AI. Building trust comes up often in our conversations with them — and in healthcare, it matters even more,” says Erik Welch, Partner, Head of Production at Godfrey Dadich Partners. That’s particularly evident in its recent video campaign for longtime GDP client InterSystems, which highlighted IntelliCare, an electronic health record with built-in AI.

The client requested a four-episode video series following a patient, Maya, through the care continuum to show how IntelliCare works and delivers better outcomes for both hospitals and patients. GDP wanted to create a series of films that would spark an emotional connection with viewers, but it was also imperative to accurately convey IntelliCare’s technology. 

To that end, GDP used medically accurate on-screen graphics and worked with an InterSystems physician on all medical details. That helped the films resonate with their intended audience: decision-makers, executives, and clinicians at hospitals around the world. The result satisfied two complex needs: it realistically showed how the technology works and also gave it a human perspective that would help drive trust with viewers.

Since the series launched, InterSystems has asked GDP for additional trailers, trade show formats, and social assets, in addition to expressing interest in a follow-up series covering more medical scenarios along the care continuum.

Serving shoppers answers they’re searching for

Nearly a year ago, OpenAI introduced ads to ChatGPT, giving an opening for retailers to target shoppers in a new way and giving shoppers the opportunity to get product recommendations that are tailored to their needs and interests. To better understand this new channel and give brands guidelines on how to succeed in this nascent space, Kepler joined Pacvue to create its Q3 2026 Commerce Media Benchmark Report, exploring exactly how people are using AI to shop and what’s working.

Working with Kepler, the clothing brand Rhone has seen early success using ChatGPT, with the platform driving 1.5 million impressions, with 92 percent of visitors new to the brand. Between June and September, Rhone saw 83 percent growth in organic AI search sessions.

The big lesson: ChatGPT ads that work are those that read like direct answers to questions, rather than a sales pitch. Pacvue’s report found that more than half of consumers use AI tools to research projects; maintaining their trust with solution-oriented suggestions is key. 

Establishing guidelines for information ownership

The rise of AI tools and opacity about the way they’re trained has heightened concerns about both data security and digital sovereignty — the ability of governments, organizations, and individuals to have control of their digital assets and shape their own future. This is a big focus for Public Digital, which works in the public sector and beyond to help organizations navigate the risks and opportunities of a fast-evolving technology landscape. 

Earlier this year, the UK House of Commons included its definition of digital sovereignty in its research on the topic, showing that Public Digital is shaping the way that the UK government approaches its work in this space.

Public Digital also published its insights on digital sovereignty in a book, Digital Sovereignty: The Power to Decide, which gives leaders clear frameworks to secure their data and establish trust in their own digital systems. The impact of that kind of work resounds.

Skepticism of AI isn’t going away on its own, but kyu companies are building trust deliberately, by testing with real people, telling the truth about what technology can do, giving users answers instead of pitches, and making sure the people who rely on systems stay in control of them.