Investing in HR technology to save on admin work means little if employees don’t trust or use it, say HR leaders
As organizations scale and their HR functions face growing pressure to operate more strategically with artificial intelligence (AI) tools, real adoption requires rethinking both process design and organizational culture before any tool goes live, according to Parvathi Subramanyam, Senior Director of Total Rewards at 1Password – a fully remote Canadian cybersecurity company. However, despite years of technology investment, human resources (HR) professionals are still spending a significant amount of their time on administrative tasks – 70 per cent of HR time in small and medium-sized businesses, according to the 2026 State of HR survey of Canadian HR professionals by HR software provider Folk and 57 per cent according to Deloitte and ServiceNow's 2024 Modernizing HR report – a burden that self-service technology was designed to reduce.
Subramanyam has watched this dynamic unfold as her organization grew from approximately 650 employees to more than 1,500 in under five years. Despite that growth, many employees and managers continue to reach out to HR directly, as if the company were still a fraction of its current size, she says.
"Some of the behaviors that we found not just from employees but actually from managers is that they tend to use HR as they would when the company was 400 people, so we've really as an HR function tried to help educate our managers, our leaders, and our employees," she says. "Managers and leaders, actually more so than our employees, to help them understand and empower them to submit requests or changes that they need to submit."
People default to direct queries
1Password introduced a ticketing system to redirect those queries, but Subramanyam is candid about the challenge of both getting people onboard with the tool and finding the balance between efficiency and keeping the people element with HR. "I think people just tend to default [to direct queries] because we have really prided ourselves in our white-glove HR," she says. "It's finding that balance between providing the white-glove service to our employees and not taking that away completely, because that's one of our values as a company – putting people first."
That tension is common across HR functions that have built their identity around accessibility and responsiveness. Understanding how employee self-service tools operate in modern HR functions is an important starting point for any team reconsidering its service model, says Subramanyam.
Design for behaviour, not just function
Subramanyam is direct about where implementations often go wrong: organizations treat self-service as an add-on rather than a redesign.
"You don't just slap on a band-aid of a self-serve tool – that’s just going to add on to the bureaucracy," she says. "You need to remap the process to make sure that it's the most efficient way possible. We're not asking people to go to multiple different platforms or to do this three times."
The reasoning is straightforward. If messaging an HR colleague on Slack takes seconds while a portal ticket requires navigating a separate system, most employees will choose the easier path every time, according to Subramanyam. Adoption follows simplicity – and that outcome is either designed or defaulted into. "You need to make it as easy as possible so that it becomes their natural behaviour as an employee population," she says.
Guardrails, training, and the AI question
Simon Blanchette, an adjunct lecturer at McGill University in Montréal whose research covers AI adoption in organizations, says the most consistent mistake he observes is deploying technology before governance and training structures are in place with clear lines on AI portal use.
"It's about purposeful planning," he says. "What’s the AI for? What kind of information are employees expected to use the AI for? And for what kind of information are they expecting to actually go and see their boss or their colleagues?"
For Subramanyam, proper assessment of the tool and training for employees is essential before rolling out an AI chatbot and self-service portal. “It will take some time and education to really make sure that whatever tool or chatbot we end up going with is really pulling from the right information,” she says. “It's not hallucinating, it's not giving our employees wrong information – we need to set the guardrails and the parameters to make sure that it's doing [its job] very well.”
Blanchette is firm that human oversight cannot be treated as optional. "The human in the loop is only as good as how well-trained the humans are," he says. "If an employee was harmed because of an AI decision [or information provided], there has to be recourse. AI can never be the final voice or the final decision-maker."
Self-service for information, not leadership replacement
Blanchette also cautions against framing AI deployment as a way to reduce leadership engagement. "Implementing AI isn’t a way for leaders to delegate their responsibilities," he says. " It might create more resources for the employees, but leaders need to still be leading in that change – to be there to support employees, create trust, and be there to actually also hear concerns."
Blanchette’s advice for organizations approaching a rollout is consistent: run a pilot phase, learn from it, then build the full implementation plan around the results.
The trust dimension carries particular weight in Canada. According to the Ipsos AI Monitor 2026, a 32-country survey, Canada topped the list of countries whose residents feel nervous about AI, with 67 per cent saying so, 17 points above the global average. As Canadian research consistently confirms, trust, not technology, is what determines AI success in organizations.
The stakes extend well beyond operational efficiency. As organizations navigate how AI can create efficiency for HR tasks and other functions, the deeper question is whether automation is building stronger cultures or eroding them.
For Blanchette, the bottom line is unchanged regardless of the technology in play. "AI can actually help you be more efficient, but organizations are still very much thriving and existing because of human," he says. "I think we want to still have a human focus and a human-centred future."
