Workers’ trust in AI: ‘Optimism tempered by caution’

Public confidence in organisations that deploy artificial intelligence (AI) is trailing far behind everyday use of the technology, according to a new report from TELUS that has direct implications for employers.

The 2026 AI Trust Atlas finds that 85% of Canadians used at least one AI‑enabled tool in the past year but only 34% of Canadians trust companies that use AI – pointing to a possible impact on AI adoption and culture and engagement in the workplace.

TELUS describes public sentiment as “optimism tempered by caution,” with 57% of Canadians believing AI can improve quality of life while also expressing concern about serious risks.

“The findings demonstrate that, despite broad familiarity with AI, trust is not a given,” write TELUS Chief Data and Trust Officer Pam Snively and President and CEO Darren Entwistle. “Public sentiment reflects real worries as well as optimism, reminding us that transparency, human oversight and accountability are not optional, but essential for confidence and stakeholder buy‑in with respect to emerging technologies.”

Trust at work is under increasing strain as AI reshapes jobs and burnout reaches record levels, according to a previous report.

Usage outpaces understanding

While AI adoption has “moved decisively into the mainstream,” understanding of the technology has not kept pace, according to TELUS’s survey of 5,487 Canadian respondents conducted in September 2025, plus qualitative polling of 1,540 Canadians in January 2026.

In Canada, 40% of respondents say they can confidently explain AI’s benefits and limitations, and 34% say they know which products and services incorporate AI. 

“Usage – 85% in Canada – far exceeds understanding, leaving a gap where people use technology even if they don’t feel confident in critically evaluating it,” the report states.

TELUS defines data and AI literacy as “the ability to create, collect, manage, evaluate and apply data and AI, in a critical and ethical manner. Put simply, it is the ability to read, write and comprehend data and AI.” A majority – 59% of Canadians – say they want to learn more about AI.

Oversight, ethics and regulation

The TELUS research shows strong expectations for human oversight and ethical safeguards. TELUS reports that 78% of Canadians say it is essential for AI systems to have human oversight, and 60% say this should apply to all AI tasks, regardless of risk level.

Another 79% agree that “the development of AI needs to incorporate ethics to minimize harms.”

Support for AI regulation is near unanimous: 90% of Canadians believe AI should be regulated, and two‑thirds (66%) say regulation is “very important.” Yet only 27% of Canadians say they are confident current laws are adequate, with just 4% “very confident.”

Most respondents believe government agencies should be primarily responsible for ensuring ethical use of AI, followed by independent regulatory bodies. Private corporations rank much lower.

Conditions for Organizational Trust

The TELUS report identifies specific practices that would increase willingness to trust companies deploying AI.

Among Canadians, 76% say their trust would increase if companies “reviewed potential harms before release” of AI systems, 73% would be more willing to trust companies that “explained AI use in an easy‑to‑understand manner,” and 69% say trust would rise if companies “asked for and listened to customers’ input regarding AI.” 

“People want to understand what’s happening, feel that technology has been built with safety top‑of‑mind and know their voices matter,” TELUS writes.

Roadmap to a ‘deeply human’ AI future

The TELUS report concludes with four recommendations for advancing what it describes as “compassionate, accountable and human” AI:

• Address the AI literacy gap through people‑first education, recognising that adoption has outpaced understanding.
• Develop AI with diverse perspectives to build systems that work equitably for all communities.
• Keep ethics and human oversight at the centre of AI decision‑making, pairing transparency with clear accountability and ongoing monitoring.
• Take a multi‑stakeholder approach to responsible innovation, bringing together government, industry, academia, civil society and communities.

“These findings offer a roadmap for industry, government and civil society toward an AI future that is not only technologically advanced but also deeply human: shaped with people, for people,” TELUS states.