USA 
United Kingdom
HR leaders at some of America’s largest companies are beginning to draw a sharper distinction between people and machines, as organizations reassess how AI agents should be deployed across the enterprise.
For the past year, companies have grappled with the idea of managing a joint workforce of human employees and AI agents that have their own names, job titles, and key performance indicators.
Although it helped early experimentation, it is now being challenged as leaders look for more tangible returns.
AI strategy shifts from agents to workflows
Speaking at the WSJ Leadership Institute’s Chief People Officer Summit, Nickle LaMoreaux, Chief Human Resources Officer at IBM, made it clear that treating AI like actual employees is not the right path forward.
“Agents shouldn’t have human names. They shouldn’t be on org charts. And they shouldn’t be given a specific job title,” LaMoreaux said.
Turning workforce data into early warnings for high-cost employees
By now, most people will have encountered a friendly chatbot or agent with focus-grouped first name during their day to day. But it seems there is a growing realization that anthropomorphizing AI can distract from where value is actually created.
“We learned this the hard way,” she said. IBM used to have a series of agents that went by names like Harry, Hermione, Charlie, and Sherlock. But it fell into a trap of focusing too much on each agent’s individual use cases rather than using them for more impactful large-scale process re-engineering.
We may be about to see a shift away from individual tools and toward system-wide redesign. Instead of asking what each agent does, the emphasis moves to how work itself is structured and delivered.
“Too many CPOs are getting so hung up on: what does this agent do, what does this AI do?” she said. The biggest bang for your buck, she said, isn’t in individual assistant-type agents that, say, help write emails. It’s in integrating AI into enterprise workflows.
“This is you declaring, top down, in the organization: here’s how employees are going to get customer support. Here’s how we’re going to run the promotion process. What’s going to be automated. That workflow, that experience, is really what we should be focused on.”
It aligns AI more closely with traditional digital transformation, rather than positioning it as a parallel workforce.
“Manage technology the way you’ve always managed technology for decades. Managing people is very different.”
HR keeps accountability with humans
While some organizations continue to experiment with the idea of AI as “digital workers,” others are taking a more measured stance.
BNY, for example, said it employs dozens of AI “digital employees” that have company logins and human managers, built on its proprietary AI platform Eliza. But even among companies embracing AI at scale, there is caution about how far that analogy should go.
At Microsoft, Chief People Officer Amy Coleman drew a clear distinction between roles and tasks, telling MSN: “I don’t really think about roles and jobs that can be automated as much as I think about tasks in a job that can be automated."
Coleman also addressed the question of whether AI could ever move into formal management structures.
“We are in that messy time of how we’re going to figure it all out,” she said.
That companies are still figuring AI usage out, merely reflects the pace of the change that is happening, but also the limits of the technology. Even as AI capabilities improve, limitations remain.
Aaron Levie, co-founder and CEO of Box, pointed to accountability as a defining boundary.
“Accountability has to lie with humans. All our laws are set up to require that.”
Governance, compliance, and performance management frameworks still rest on human responsibility, regardless of how much work is automated. At the same time, the scale of transformation is significant.
“This is the biggest shift we’ve ever seen in corporate work,” Levie said.
The challenge now is moving beyond experimentation toward integration. Treating AI agents as people may have provided a useful entry point, but leaders increasingly suggest it risks limiting the technology’s impact.
The emerging consensus is that AI should be embedded into workflows, not given names and positioned alongside employees.
