Trust and Generative Artificial Intelligence (AI) is a Complicated Relationship

Bruce MacLellan

In our decade of studying trust in Canada, few topics bring more complexity than with generative Artificial Intelligence. It already seems clear that trust will not come easily to this sector, and that communications people have an important role to play.

Off the top, anything with artificial in its name presents an obvious trust challenge. The word traces back to the Latin word artificium, which in turn has roots in the word ars, meaning art. We all know art is subjective and open to expression and interpretation. Is this the future?

An irony in this trust discussion is that gen AI is already embedded in our lives. The 90 per cent of Canadians with a smart phone are using AI daily. The consulting firm Accenture offers an optimistic forecast that “gen AI is poised to provide the most significant economic uplift and change to work since the agricultural and industrial revolutions.”

I am ready to share this optimism, although something that dominates an entire era will follow an unpredictable timeline. Contextual factors will also have an influence, such as the prevalence of freedom or dictatorship, economic prosperity, and crises caused by events in nature or human conflict. There will be known and unknown risks that impact our relationship with AI.

Trust is defined as the willingness to be vulnerable in return for the expectation of positive benefit. In public understanding of gen AI, these are early days. Unfortunately, the positives are not yet well communicated.

At this initial stage of the AI timeline, much of the discussion is about guidelines and guardrails. While appropriate, the strong emphasis on risks makes it hard for people to appreciate the rewards.  The past six years have been busy for AI, yet our CanTrust Index research has found a decline in trust in AI to contribute to the Canadian economy. On this question, trust fell to only 33 per cent in 2024, down from 39 per cent in 2018.

As we enter the next phase of AI, the lift in trust will only happen if each sector communicates the positives. Currently, our research finds trust in benefits is tepid. Only a third of Canadians expect AI to improve the consumer experience. Only 27 per cent expect AI to support effectiveness in the healthcare sector. It will take years, but trust can be built when safety and benefits are explained.

We know that trust in AI is higher among younger people and post-secondary education graduates. As generations change, comfort with technology in general should increase. In the meantime, if you use AI, talk about AI. Explain the safety and the benefits. Teams and responsibilities should be assigned to communicate about gen AI. Customer understanding of benefits should be tracked. There is a massive amount of literature available to help. The Institute for Public Relations has very recently published a report on AI in Organizations, with recommendations from communication leaders.

Communicators have a major role to play in ensuring that gen AI is understood and seen as beneficial. AI can help. Let the creativity begin.