Last week, TAP Network members joined us for a timely conversation on AI, privacy, and legal risk in the workplace, featuring insights from Jordan Michaux and Justin Wong of Roper Greyell LLP.
As AI tools become deeply embedded in recruitment, people management, and day-to-day operations, one thing became clear: this is no longer a future issue — it’s a governance issue right now.
AI in Hiring: Efficiency Meets Expectation
While AI is widely used for tasks like résumé drafting, mock interviews, and interview note-taking, both candidates and employers continue to expect human judgment at critical decision points.
Survey data shared during the session showed that 87% of candidates prefer a person conduct their initial interview, and 85% want their résumé reviewed by a human. Employers echoed this view, noting that human input remains essential, particularly when assessing soft skills, judgment, and cultural alignment. Neither group expressed confidence that AI can effectively evaluate these dimensions on its own.
That expectation isn’t just cultural, it’s legal. Bias embedded in AI systems, reliability concerns, and the difficulty of explaining AI-driven decisions create real exposure under human rights and employment legislation. High-profile examples, from abandoned AI hiring tools to documented bias in generative AI systems, show how quickly things can unravel without clear oversight.
The takeaway for tech leaders: AI can enhance hiring processes, but accountability and defensibility still rest with people.
Privacy: The Hidden Risk Multiplier
A major theme throughout the session was privacy risk, particularly as AI tools dramatically increase the amount of personal information being collected, stored, and analyzed — often automatically and invisibly.
Canada’s privacy laws were not designed with generative AI in mind, creating a grey zone for employers. Privacy commissioners have raised concerns about transparency, consent, accuracy, and the inability to track how personal information is used once it enters an AI system. As the panel emphasized: more data can mean more problems, especially in remote and distributed work environments.
Importantly, organizations remain accountable for privacy breaches, even when AI tools or third parties are involved.
Reliability, Hallucinations, and “Agentic” AI
Beyond privacy, the speakers explored growing concerns around AI reliability and hallucinations. Even advanced, domain-specific AI tools can produce confidently wrong outputs — a serious issue when those outputs inform hiring, performance, or termination decisions.
As AI systems become more autonomous (“agentic AI”), the legal and reputational stakes only increase. Recent legal cases show that courts are holding organizations responsible for AI-generated errors — the technology itself is not a shield.
What Employers Should Be Doing Now
Rather than trying to solve this purely with technology, the panel stressed that policy, people, and process matter most. Practical risk mitigation includes:
- Clear AI and privacy policies
- Defined rules around authorized AI use
- Ongoing employee training
- Human oversight and build redundancy in decision-making
- Monitoring, enforcement, and iteration as tools evolve
AI adoption isn’t an “if” or a “when” — it’s a “how.” And getting that “how” right now will matter even more as regulation continues to catch up.
Key Takeaways for Tech Employers
- AI doesn’t remove employer liability. Organizations remain accountable for decisions influenced or made by AI, including privacy breaches, bias, and unreliable outputs.
- Human oversight is essential. Employers and employees alike still expect people to make — and be able to explain — hiring and employment decisions.
- Privacy risk increases quietly. AI tools can dramatically expand the collection and use of personal information, often without clear consent or transparency.
- Bias and reliability remain unresolved. AI systems can embed bias and hallucinate with confidence, creating real exposure under human rights and employment law.
- Policy beats tools. The most effective risk mitigation starts with clear AI and privacy policies, defined rules of use, training, and enforcement, not just technology.
- AI adoption is a “how,” not an “if.” Employers should take an intentional, iterative approach to AI integration as regulation continues to evolve.
Looking to Strengthen Your Digital Leadership?
As AI adoption accelerates, the real challenge isn’t just implementation — it’s governance, risk management, and leading change effectively across your organization.
TAP Network is proud to partner with Simon Fraser University on the Digital Transformation Management program, designed to equip leaders with practical frameworks to navigate emerging technologies, including AI, with clarity and confidence.
If you’re thinking beyond tools and toward long-term capability building, this program may be worth exploring.
👉 Learn more about the Digital Transformation Management program
https://tapnetwork.ca/lp/digital-transformation-management/
