ClickUp's 3-to-1 agent-to-employee ratio puts culture, trust, and belonging at risk, say experts
When ClickUp chief executive Zeb Evans told employees earlier this year that they should work through an artificial intelligence (AI) agent trained to act like him before reaching out directly, the productivity software company's staff encountered a fundamental shift in what it means to have a manager.
The San Diego-based firm, valued at approximately $4 billion, now runs roughly 3,000 internal AI agents — a three-to-one ratio of agents to humans, following a round of layoffs that reduced staff from 1,300 to 1,000 — with every workflow and task touching one in some form. It is, by most measures, an aggressive documented deployment of agentic AI — AI systems capable of autonomously executing multi-step tasks. For HR leaders in Canada, it raises a question that goes well beyond efficiency metrics: what happens to organizational culture and leadership structure when AI becomes the first point of human contact?
The business logic isn’t hard to follow. AI agents operate around the clock, process requests quickly, and keep costs low. Evans himself no longer checks email, company chat threads, or project dashboards. An agent scours all of it on his behalf, distills what he needs to know, and delivers it in a format resembling a newspaper briefing, he told Fortune. For managers and HR leaders drowning in routine queries, there’s obvious appeal to that approach.
Trust, connections with leadership getting watered down
But the HR community is watching cautiously — and not without reason. Stacy Parker, co-founder and Managing Director of Blu Ivy Group in Toronto, says that trust in leadership is key to organizational success, and inserting AI agents for some leadership tasks is risky.
“The last few years have already gotten watered down with respect to what leaders have had to navigate since COVID — it's been constant change at a pace that’s unprecedented, but that constant change has really muddied strategic messaging and leadership trust,” says Parker. “Now we've added AI, which is already industry-wide and not trusted, and we're asking teams to use it and we're starting to communicate using it internally, so the nuance of relationship is getting watered down so that that human connection — which is a part of what makes work work — is being eliminated.”
Parker sees a convergence of troubling conditions when AI becomes the first contact for employees for any issue. “You have low trust in leaders, low trust in AI overall, and low psychological safety — all merging and now creating distance and additional skepticism,” she says.
A 2025 global poll by Workday, AI Agents Are Here – But Don’t Call Them Boss, found that 82 per cent of organizations are actively deploying AI agents, with 88 per cent believing agents will reduce workloads. While 75 per cent of employees said they’re comfortable with AI agents working alongside them, only 30 per cent said they’re comfortable being managed by those agents.
What goes quiet before it shows up in data
The gap between organizational enthusiasm and employee acceptance is one that HR leaders can’t afford to overlook, according to Simon Blanchette, faculty lecturer at McGill University in Montréal and a change management consultant. Blanchette sees the ClickUp model as a stress test — one whose full consequences will only become clear over time.
“When we look at this specific case, time will tell if they succeed or not,” he says. “We're still very much in the early days of AI, and if you compare it to things like electricity, for example, we're still very much in the infancy of it.”
Blanchette’s deeper concern is what gets lost before leaders notice. He draws a parallel to the pandemic, when new hires who never experienced a physical workplace struggled to build loyalty to something they could not feel.
“If you don't have human connections — working with colleagues, mentoring, relationships — what are you becoming loyal to?” he says. “The notion of belonging becomes really abstract if it's only you on a computer with an AI agent as your main colleague and supervisor.”
Compounding isolation, culture erosion
The compounding effect matters too. A 2024 Gallup survey found that nearly half of remote employees report feeling lonely always or sometimes while working — and that data predates the widespread rollout of agentic tools. Layering AI-first communication onto an already-fragmented hybrid workforce risks accelerating an erosion that is already underway, according to Blanchette. Canadian HR leaders tracking the rapid rise of agentic AI in Canadian workplaces will find the warning signs aren’t hypothetical.
Parker is equally pointed about the organizational cost, based on what she’s seen in organizations. “When you remove management from collaboration and mentorship, you lose clarity of what the objectives are for execution and you reduce your reputation — not just as an organization, but as a CEO,” she says. “Your reputation starts to slide and you become entirely reliant on AI or technology to convey the messaging.”
Parker and Blanchette agree on a common challenge for organizations: the costs of cultural erosion are real but slow to register in traditional metrics. Turnover and engagement scores are lagging indicators — by the time they move, the damage is already done.
Leadership and employee experience shape customer growth
Parker advocates for what she calls the human value chain: a framework that connects leadership reputation and employee experience directly to customer retention and revenue indicators. “Employees that are engaged perform measurably better on the stock exchange — but at this moment in time, that's not enough," she says. “We have to be looking at how culture and people are shaping customer retention and customer growth.”
Blanchette takes issue with how organizations are defining productivity in the first place. An AI agent that frees up an hour of an employee's day only creates value if that hour is reinvested strategically, he says. “More work is not always more productive or more value-added,” he says. “If I get an extra hour, where can I actually invest it to create more value that I wouldn’t have been able to do otherwise?”
HR leaders navigating the top legal and operational risks posed by AI agents in 2025 and beyond will find that governance is as much about people as it is about process.
Getting the balance right
Neither expert argues against agentic AI adoption, and they agree that the question is how and where to draw the line. “Globally, to be competitive, we need to start implementing and adopting AI more, but at the same time we need to do it in an inclusive and responsible way,” says Blanchette. “And also, we need to improve organizations’ digital maturity, because currently what we're seeing is that even if some firm wants to implement AI, they don't necessarily have all of the infrastructure in place.”
Blanchette recommends a pilot phase before any full rollout — with a clear post-mortem — followed by explicit guardrails defining what AI is for and what it isn’t for. He’s direct about the risk of skipping the human infrastructure step. “Implementing AI isn’t a way for leaders to delegate their responsibilities,” he says. “Leaders really need to still be leading in that change, to support employees, create trust, and actually hear concerns.”
Parker sees AI as most valuable when it handles task management, administrative co-ordination, and structured onboarding milestones — leaving the relational work to people. She draws on the return-to-office debate as a reference point: “We already had that tested over the last few years with remote work,” she says. “That was part of the great reasoning for why we were bringing people back into the office — the loss of insight gathered sitting side-by-side.”
The ClickUp model may ultimately prove workable, both say, if it’s built around emotional intelligence rather than in spite of it. Parker's vision of effective agentic design includes AI that asks how an employee is feeling, what they are struggling with, and — crucially — when they would benefit from speaking to a human leader instead.
“People always buy and stay based on relationships, not on product features,” Parker says. "And the more you remove that relationship, you’re removing your message of who you are and what you stand for.”
