Consultants blame rollout design for the 3 in 10 workers undermining AI strategy

The surveys measure how common the behaviour is. Writer and Workplace Intelligence's 2026 AI Adoption Survey counts three in 10 employees as actively sabotaging their company's AI strategy ^[3], about 720 of the 2,400 respondents ^[8]. The count bundles four behaviours: declining to use the tools, disregarding the guidelines, deliberately producing poor-quality outputs, and putting proprietary information into platforms that have not been approved ^[5]. The article does not say whether either survey asked respondents why they acted as they did.

The explanation comes from two consultants. Brent Dul, executive vice-president at Randstad Canada in Toronto, says organizations deploy AI without first investing in the people expected to use it ^[9]. He treats the sabotage figure as a trust problem. "That's just indicative of employees not being able to buy into AI because they feel it's designed to replace them," Dul said ^[10]. He also said confident Canadians face "an operational reality gap in how they see AI as impacting their future career" ^[6]. The TD figures fit that picture, since acceptance of AI is high in the report ^[4] while only a minority say their employer's training is adequate ^[2].

Aram Arslanian, president and CEO of Cadence Leadership + Communication in Vancouver, puts the fault in sequencing. He cites the stages of change model, which runs from pre-contemplation through contemplation, preparation and action to maintenance, and says most rollouts ignore it ^[11]. "Individuals and companies are skipping those first two crucial steps and then just trying to hop right into training or right into 'This is what we're going to do,' and your audience isn't there," Arslanian said ^[12]. He estimates that neutrals, who watch and take their cues from those around them, make up about 60 per cent of any workforce ^[14]. "Neutral people are almost what I consider to be like swing voters in an election," he said ^[15].

Arslanian himself names where the training explanation stops working. Some challengers are detractors, and of those he said "they probably shouldn't be in the organization" ^[16]. He also called passive resistance, where employees appear to comply while undermining the initiative anyway, one of the most damaging forms and the easiest to miss ^[13].

We think the training explanation is the better-supported one for most workers, because the TD report shows acceptance and a training shortfall together. It remains an inference. It rests on two consultants and two surveys that measured different things: skill overstatement among Canadian workers in one ^[1], sabotage across five countries in the other ^[3]. Whichever cause dominates, leaders start from poor odds. Kearney's Transformation Study 2026 found that only 29 per cent of corporate transformations deliver their intended value consistently ^[7].