Workers are fine with AI agents as assistants—but draw a firm line at letting them take charge.
A global Workday survey shows optimism around AI agents is high—until you put them in charge.
Employees trust AI to help, not to lead
AI agents are here. According to a global survey from Workday, 82% of organizations are expanding their use of agents, driven by expectations of reduced workload, faster innovation, and better productivity.
But there’s a hard stop: only 30% of respondents are comfortable being managed by an AI agent. And just 24% are okay with AI operating in the background without their knowledge. It’s a clear boundary. Workers want AI to assist, not command.
This isn’t Luddite paranoia. It’s basic sense. AI agents can write simple code, analyze data, and suggest smarter workflows. But they can’t mentor, mediate conflict, or make high-context judgment calls. Trying to replace human managers with bots isn’t just bad optics—it’s operationally dumb.
Let people lead, let agents execute
The Workday report nails the nuance: trust grows with exposure. Teams that work directly with AI agents build confidence over time. 83% of employees trust their organizations to use AI responsibly, and 75% feel comfortable using AI for tasks like upskilling and project support.
But that trust tanks the moment AI crosses into authoritative territory:
- 🔢 75% are fine with agents assisting
- ⛔ Only 30% are okay being managed by one
- ❓ Just 40% want agents making financial decisions
Leaders pushing AI into management roles are missing the point. This is a moment for human-led, AI-accelerated operations. As we outlined in our piece on small team advantages, nimble ecommerce operators are winning because they’re using AI to speed up execution, not to automate authority.
Big companies are overstepping
Here’s where the cracks are showing. Enterprise orgs see AI as a governance tool—a way to scale decision-making without more managers. But that move is premature. 48% of Workday respondents fear AI will just increase pressure to perform, not reduce workload. And a third are seriously worried about erosion in critical thinking and human interaction.
That fear is well placed. Most enterprise AI rollouts are led by IT and procurement teams that prioritize scale, not user trust. And while some companies are launching agents that streamline ops—like Walmart’s persona-focused super agents—others are pushing agent authority before they’ve earned team buy-in.
Even in finance, where AI is set to help plug staffing shortages, 76% of professionals see agents as support tools, not as compliance leads. Trust lives in execution, not in governance.
The real ROI is speed, not hierarchy
Forget the AI org chart reorgs. The best use of AI agents right now is tactical:
- ✅ Speeding up weekly reports
- ✅ Surfacing personalized product recommendations
- ✅ Automating low-risk workflows
- ✅ Enhancing onboarding and internal comms
Companies looking to extract ROI from agents need to stop fantasizing about fully autonomous teams and start investing in tools that make their best people faster.
This isn’t about replacing managers. It’s about unblocking them.
So what?
AI agents are here to stay. But turning them into middle managers is a fast track to employee distrust and failed rollouts. The smarter move? Equip your team with tools that help them work faster, think bigger, and automate the junk.
Save the AI boss fantasy for your keynote slides. In the real world, humans lead. Agents assist. That’s the only way this works—at least for now.
The Weekly Rundown for Ecommerce Insiders
