AI-driven workforce data is changing talent decisions, but HR must still overcome concerns over privacy, data quality
People analytics powered by artificial intelligence (AI) can now predict attrition, identify high-potential employees, and synthesize engagement results in minutes. Many Canadian organizations, however, are still to act: As of mid-2026, 19.2 per cent of Canadian businesses were actively using AI, according to Statistics Canada’s Analysis on artificial intelligence use by businesses in Canada – although this figure tripled in two years, it’s still well below rates in comparable economies. Yet within human resources (HR) teams, the appetite tells a different story, according to Sheila Thomson, HR Director at EastGen in Guelph, Ont.
"All of the people in my human resources (HR) network, we are high users of it,” says Thomson. “I'm seeing HR leaders and HR teams wanting to adopt it as much as they possibly can.”
Where Thomson sees some hesitation is in data confidentiality – feeding sensitive employee information into AI platforms carries real risk in a tightly regulated environment. "I'm always hesitant about confidentiality or results that are worded in a way that can put us at risk – because HR is highly governed and legislated, a word can have a significant difference," she says, pointing to the legal weight that distinguishes terms like "may" and "will" in employment documentation.
Governance is the defining challenge for AI-powered data analytics in HR, according to Permpreet Soomal, Vice President, Human Resources at MDA Space in Toronto. "When we're thinking about the type of data that we're working with, how do we protect that and make sure that our employees understand the ways to protect our IP, and we're using it responsibly,” says Soomal. “It continues to be one of the challenges for how we use it effectively."
From dashboards to predictive people analytics
Both Thomson and Soomal identify the same fundamental shift in what AI-powered people analytics tools available to Canadian HR teams can deliver.
"The biggest shift that I'm seeing in leveraging AI to make more data-informed decision-making is really the shift from looking-back reporting to more forward-looking predictive insights," Soomal says. "Asking ourselves the questions, not necessarily the ‘what happened,’ but instead ‘what will happen’ and how do we optimize this from a business outcome standpoint."
At MDA Space, that means applying machine learning to identifying high-potential employees rather than relying on gut-feel nominations, according to Soomal. "We're looking at how do we do that in less of a subjective gut-feel nominations type of process and using more machine learning to analyze success criteria like performance trajectory, skills acquisition rates of our employees, and any project complexities that they've mastered or influence that they have within their networks across the organization," she says.
In employee engagement analysis, Thomson has found the strongest returns on the shift to data-driven HR decision-making in Canada. "I would spend months trying to pull together the analysis, the data, the review, and the comments, but now I can do it with a click, and that’s the efficiency that I love," she says. She cautions that the tools demand literacy, not just access. "If you're just jumping into only using it, then I think it can set you up for inefficiencies in the long run because it might not be accurate if you're not prompting it properly.”
Navigating privacy without losing workforce data insight
Employee privacy remains one of the most persistent barriers to AI-powered people analytics in Canada. Provincial and federal legislation strictly governs how data is collected – and AI adds new complexity to a demanding compliance environment.
Soomal advocates working at the aggregate level, analyzing patterns and themes rather than feeding personally identifiable information into AI models. "Using AI to answer some of those problem statements that are more thematic at the aggregate level is a great way to leverage insights without getting into personal identifiable information feeding some of these models and creating risk around data privacy," she says.
That approach helped Soomal’s organization address a significant strategic risk, she says. By mapping retirement risk across key job families over a five-year horizon, the team identified niche skill shortages and deployed AI to capture institutional knowledge before it walked out the door. "Using AI tools to capture, synthesize, and ingest some of that unstructured technical experience that we captured from our individuals – using that to do everything from process mapping, interdependent workflows, or capturing best practices to understanding their mindset to approaching complex problem solving," Soomal says.
Thomson says she strips out names and identifying details before any data goes into an AI platform. "I typically take out names and anything that would identify someone if I'm putting data in for it to analyze – just to try and do an extra level of confidentiality," she says.
Making the business case beyond the dashboard
Only three per cent of C-suite executives identified HR as the function where AI is most driving decision-making, according to the Forbes Research 2025 AI Survey – a gap both Soomal and Thomson see as important to close.
Soomal says it’s key for HR leaders to connect data to what executives really care about. "The first thing that we have to do with people analytics is connect insights to business outcomes, and the best way to do that is to really speak the language of the business," she says, adding that the days of HR leaders showing up just with a metrics dashboard are gone. "It’s more of how are we’re using people analytics to make better business decisions, how efficient our business is, where we're removing risk, or how we're growing the business."
Thomson draws on experience in civil engineering, where numbers were required at every meeting. "Using numbers has been ingrained in me from an earlier stage in my career – the business needs to appreciate and want the metrics, but then also see the value of the efficiency of using AI," she says.
Building a stronger business case for AI investment in HR starts with rethinking how data is shared, according to Soomal. "Moving from purely HR data that’s confidential to how can we use that data in a safe way for managers to glean insights and lead their teams better in a democratized way is a step that we need to take, if we're going to put the power in the hands of the leaders that can actually influence higher performance through behaviour change in their teams," she says.
A key to fully using the wealth of data analytics now available to HR leaders thanks to AI is overcoming the fear of getting started without perfect data and getting creative, according to Soomal. “Data is never perfect, so waiting for that flawless dataset before you get started, I think, is a really big risk in my mind, especially when things are moving so quickly,” she says. “The risk of not using some of these tools to identify trends or directionally give insights into people analytics feels like a big miss at an organizational level.”
This article is part of our Monthly Spotlight series, which in August focuses on AI in HR. Full coverage can be found here.
