The Green Room podcast
Can we trust something we don’t fully understand?
Trust is both delicate and hard earned. But it can also be the difference between experimentation and scale.
Is trust the biggest question for AI?
It’s not just “what can it do?”, but “what confidence do we have in the systems behind it?”
AI systems are becoming more capable – but capability alone doesn’t determine whether people use them at scale. It requires trust for people to use them for important decisions or for whether organisations will roll them out widely.
Trust is shaped by more than accuracy. It includes how a system uses data, whether decisions can be understood, how risks are governed and what happens when something goes wrong. But it’s also contextual – the confidence we need for a low-risk experiment is very different from the confidence we need for sensitive, business-critical use.
In this episode, Avtar Benning and Simon McDougall explore why trustworthy AI is not a constraint on innovation, but one of the conditions needed to move from experimentation to scale.
Watch below, or listen on Apple Podcasts, Spotify, YouTube or wherever you get your podcasts. Read the transcript here.
AI has become distributed and widely used very quickly. We haven’t had time to build all of these components up and we’re going to have to build some of them as we go along.
Simon McDougall – Chief Strategist for AI & Privacy, ZoomInfo
“You don't understand how a jet engine works, but you're confident enough to get on an aeroplane.”
There’s plenty of things in life that we trust without fully understanding how they work – but is AI one of them?
The question of trust isn’t new, but the technology is. And as AI has an ever-increasing influence on how we work and live, understanding it feels more important.
But - as Avtar Benning, director in Deloitte’s Trustworthy AI offering, and Simon McDougall, Chief Strategist for Privacy & AI at ZoomInfo, agree - trust isn’t about understanding every individual component. Instead, it’s built through confidence in the systems around it - reliability, testing, governance and experience. The challenge is that AI’s fast pace of adoption means we’re building some of those components as we go along.
Listen to Simon explain the story of trust through a story about the White House in 1891.
Want to hear more about trust? Listen to our episode, “What does it take to build trust?” now.
Trust is more important than more intelligent models
Avtar Benning argues that trust is the real barrier facing firms right now when it comes from moving AI into full scale, enterprise-wide adoption.
Many people will try an AI tool before they fully trust it. Organisations may also run pilots and proofs of concepts – but important questions can remain unanswered.
When a system becomes part of everyday work, handles sensitive information, influences high-impact decisions, or has autonomous space to operate in, the threshold of trust changes.
People need confidence that the system will behave as expected. That someone remains accountable. That each decision it makes, even when it’s wrong, can be understood.
Without that confidence, organisations remain trapped in experimentation, even if the capability is ready to be scaled. So, trust in our AI systems, built by assurance, is more critical for organisations than investment in more intelligent models.
Trust isn’t built equally – but are we doing enough?
What does it take to create systems where decisions can be understood, measured and trusted? Because trust isn’t something technology can ask for, it’s built through experience, confidence and fairness.
AI isn’t just making us more productive. It’s influencing decisions that impact people’s lives and driving widespread change in how we work, think and choose. If the future is AI-led – understanding whether we can trust what’s inside the “black box” matters more than ever.
Watch below, or listen on Apple Podcasts, Spotify or wherever you get your podcasts. Read the transcript here.
I think what's more important than whether it sits with any particular
individual is having clear allocation of responsibilities across the board.
Simon McDougall – Chief Strategist for AI and Privacy, ZoomInfo
Let’s make this work.
To view this video, change your targeting/advertising cookie settings.
