Leaders from Microsoft and OpenAI Weigh in on Why Trust in AI is a New Competitive Advantage
Many Americans are familiar with AI — but that doesn’t mean they trust businesses to use it responsibly. In fact, just 27% of Americans said they do, according to a newly released Bentley-Gallup survey. On October 8, Bentley asked some of the biggest players in the industry, including Microsoft and OpenAI, about the factors influencing mistrust, and what businesses can do about it.
Gallup CEO Jon Clifton introduced the report findings as business leaders gathered in the nation’s capital to explore how businesses can leverage AI while also building public trust in how the technology is used in advertising and the workplace.
The event — The Tech Trust Gap: As Skepticism Rises, How Should Business Embrace AI? — featured a conversation with Ginny Badanes, general manager of Tech for Society at Microsoft, and Felipe Millon, head of Government Go to Market at OpenAI. The panel was followed by a "fireside" chat with Rumman Chowdhury, founder of Humane Intelligence Public Benefit Corporation. Chowdhury also served as the first United States Science Envoy for Artificial Intelligence under the Biden administration.
Key Takeaways
Companies successfully adopting AI combine experimentation with appropriate guardrails
Trust in AI is built over time through practical use to understand and demystify it
Business and government need to collaborate on best practices around the use of AI to take the burden off consumers
AI developers need to be transparent about their intentions and design models to behave in ways that align with those intentions
Closing the Gap Between AI Adoption and Public Trust
Millon believes that experimentation is the way to navigate the “relentless” pace of the AI industry. “Trust is earned, not given. Where we see trust build is when people have started adoption to really understand it. To learn AI, you have to use it.”
He likened it to reading about fitness versus going to the gym. “Seeing AI for yourself — both the amazing things and where this technology falls flat — it starts to demystify it. There’s an awe of AI that almost feels magic. Using it over time lets it get less of the ‘magic’ and more of the practical application.”
Millon suggested reframing messaging from the human versus the machine to the human and the machine — and then asking what that looks like. From the perspective of a company that builds AI, he said, “The responsibility on the developers of AI is to be transparent on how we intend our model to behave, and then aligning our model such that it behaves in that way.”
Leaning into Disruption
Badanes called for industry to shift the burden of trust from consumers to businesses by coming together with initiatives around labeling and consumer information, for example. “Consumers are not forensic scientists.”
Regarding workers trusting companies to use AI responsibly, Badanes said, “Any time a new technology that is as transformational as AI comes into place, there’s always disruption. That disruption could be that you don’t have certain jobs anymore because the technology has replaced the need for that, but what always happens is that new jobs are created or you need more of the jobs that exist.”
History, she added, can be a guide for the future of work. “Look at the points where we’ve had these transformational technologies take over: What has it looked like? How has the economy adjusted? It isn’t going to happen automatically. There are responsibilities from the tech companies to make sure we are putting out the right skilling and training — working with governments and communities so that they’re seeing the benefits of this and are able to meet the moment with the right skill set.”
AI Optimism, and What’s Missing
The program included a fireside chat with Rumman Chowdhury, founder of Humane Intelligence Public Benefit Corporation, an organization built around independent community-driven testing of AI systems. Bentley Associate Professor of Mathematical Sciences Noah Giansiracusa led the conversation.
“AI doesn’t do anything — people do stuff,” Chowdhury said of anthropomorphic messaging about AI as a scheming actor. “We make [AI] seem human. That’s the core of the trust paradox. As people get more used to using AI, they realize it’s not this magical, fantastical technology; it is just another human technology.”
She called for clear guidance and regulations, as well as hiring employees who understand how people, society and technology interact.
“What’s missing in the AI conversation are all the human things it might enable,” Chowdhury said. “Human beings are fundamentally very resilient and very capable of making really good things. Sometimes the anthropomorphic narrative gets in the way of doing that because it’s disempowering the things we do best, which is build and solve problems and work with each other.”
It was a packed house as more than 120 participants gathered to exchange ideas and collaborate.
Located in the Penn Quarter in the heart of Washington, D.C., Gallup’s global headquarters is just a short walk from the National Mall and houses the Grand Hall of the former Masonic Temple.
Why Skepticism About AI is Good
The panelists emphasized that addressing public skepticism about AI starts with listening to people’s concerns and showing them how it can improve their lives. “Skepticism from young people will be a strength for us in the technology industry because they are pointing out things that we are maybe missing,” Badanes said. “Their future depends on us doing this right.”