Follow ZDNET: Add us as a preferred source on Google.
At the Semafor World Economy summit this week, the question dominating nearly every panel was the same: will AI eliminate jobs? Most CEOs fell back on the familiar augmentation argument — that AI amplifies what workers can do rather than replacing them. But the data tells a more complicated story. Entry-level positions are shrinking, and AI is on track to compete with humans on a wide range of tasks as early as 2029.
Regulation is one response. But the more immediate lever may be workforce upskilling — the question is where that's actually happening versus where it remains a talking point.
Also: This AI expert says the job apocalypse isn't coming, even if you're a coder — here's why
I sat down with several leaders at the summit to understand what their companies are doing to prepare employees for a materially different near future — and what approaches are actually working.
Without corporate initiative, workers are largely on their own. On the legislative front, the bipartisan AI Workforce Training Act, introduced in February, would offer tax credits to companies that train employees in prompt engineering, data literacy, machine learning, AI ethics, and related disciplines. The Trump administration's latest AI regulatory framework also calls for AI training and apprenticeship programs.
But until policy catches up, the heaviest responsibility falls on employers themselves. A Gallup poll published Monday found that manager support is the single biggest driver of successful AI adoption at work.
Also: Half of all US employees use AI at work now — and waste almost 8 hours a week doing it
Dan Priest, chief AI officer at PwC, works with clients across industries on AI strategy and has seen the full spectrum of upskilling approaches — from highly structured programs to largely informal ones. Whatever the format, he sees meaningful workforce development as a baseline expectation of competent leadership, not an optional add-on.
One example: PwC helped Wyndham Hotels build an agentic system to handle customer requests, cutting average call times by at least 30%. Rather than shrinking headcount, Wyndham redirected freed-up time toward training employees in higher-value skills — like delivering a more personalized, engaged guest experience.
"The objective wasn't to replace those people," Priest emphasized. He credited Wyndham's reinvestment in its workforce as central to the program's success.
Also: Turn AI chaos into a career opportunity by preparing for these 4 scenarios
He oversaw a similar initiative at Lucid Motors, where AI-assisted improvements to financial forecasting tools led directly to new skill development — not to headcount reductions or reduced reliance on the employees involved.
Cisco takes a more mandated approach. EVP and chief customer experience officer Liz Centoni said AI training is a companywide requirement — not a suggestion.
"It's a requirement for everyone across the organization because there needs to be at least a core understanding of AI," she said. With 98% of Cisco employees using AI tools daily, the company's learning infrastructure — long established for both staff and clients — has become the foundation for a hands-on AI training program complete with a tiered completion system modeled after martial arts: blue, white, and green belt certifications for different modules.
But Centoni emphasized that Cisco's ambitions go beyond layering AI onto existing roles. The real question, she said, is more fundamental.
"More than just bolting AI onto existing workflows, how does the work itself need to change?" she said.
Also: AI is more likely to transform your job than replace it, Indeed finds
Mihir Shukla, CEO of Automation Anywhere, echoed that framing. Effective upskilling, he argued, requires a clear strategy for how work should actually get done — not just a toolkit distributed to employees. That means designing for "autonomous IT, autonomous supply chain, autonomous finance, autonomous claims, autonomous healthcare" at a systems level, rather than "sprinkling AI into the workforce." The most successful programs, in his view, let employees learn by working alongside real agents in live workflows.
"We recently challenged our engineering team to create software that was fully autonomous end-to-end, with no human coding or any employee who touched it," Shukla said.
Back at PwC, Priest has found that internal upskilling succeeds when it resists the one-size-fits-all temptation. Generational differences, in particular, require different approaches.
"Different generations in the workforce are going to respond differently to AI," he said. "We made short video explainers for specific tasks, and younger hires respond really well to that format. But that doesn't work for partners. For them, we bring them into a room and discuss how to develop their non-technical skills based on what AI is handling now."
Younger employees, Priest said, generally show strong enthusiasm for adopting AI. But leaders need to be thoughtful about those who don't — particularly experienced specialists who've spent decades developing deep expertise.
Also: Nervous about the job market? 5 ways to stand out in the age of AI
"You've got people who've worked for 20 years to become specialists in their fields, and you're telling them they have to change everything they've been doing," he said. "That's extremely hard. Leadership has to clarify what's changing — but they also need to say what's staying the same."
His practical advice: companies should "pick their spot" — prioritizing the employee groups where AI implementation is most urgent relative to their business objectives — rather than trying to upskill everyone simultaneously. He's not worried this creates lasting blind spots.
"I haven't seen a category of job that's being ignored," he said, noting that gaps in coverage tend to be temporary, not structural.
Centoni said Cisco applies a similarly role-specific logic. Course offerings are calibrated by job category, with depth of training matched to the complexity of AI use cases in each role.
One unexpected byproduct of this targeted approach: AI has surfaced institutional knowledge that had long been buried under repetitive, automatable work.
"Problems you cannot prompt your way through — that knowledge is just sitting in an engineer's head," she said. "When we used AI to automate all of the things that could be automated, that elevates these folks."
That realization is now reshaping how Centoni thinks about future hiring — both the kinds of people Cisco seeks and the metrics used to evaluate their success once onboarded. She doesn't have all the answers yet, but the question is clearly top of mind.
Also: Is AI coming for your job? Here's one labor indicator that could soothe your fears
Centoni stopped short of claiming that upskilling is directly preventing layoffs at Cisco. Shukla took a similar tone — but both made clear that AI development and career development are, in their frameworks, inseparable.
"Every employee needs to operate in a place where you push a model to where it fails, and then you understand where your unique value lies," Shukla said. "For us, AI upskilling is directly tied to career growth. Whether you stay at our company for decades or a few years, we feel a responsibility to cultivate a mindset around leveraging AI and mastering how to use it."
Priest shares that orientation toward talent, which is why he's skeptical of aggressive AI-driven layoffs as a strategy. The companies actually getting results, he said, are leading with upskilling.
"It all centers around talent," he said. "There's a reason there's a talent war in Silicon Valley right now." Shukla agreed that cutting too fast destroys institutional knowledge — and often means companies must rebuild roles at greater cost later.
That's a meaningful counterpoint given how aggressively some companies have moved. Anthropic CEO Dario Amodei recently escalated his predictions about AI-driven job displacement. This week alone, Snap laid off 1,000 employees, partly citing AI capabilities. And over the past several months, Meta, Oracle, Block, and others have made substantial cuts. Whether these moves reflect genuine AI-driven efficiency or simply correct for earlier over-hiring is hard to determine from the outside — but the rhetoric risks encouraging less AI-sophisticated sectors to follow suit without the same due diligence.
Priest also flagged a practical argument for keeping humans in the loop: accountability. "We're a regulated business," he said. "Something goes wrong — you don't tell your agent, 'Hey, you're in trouble, you did a bad job,' right? It's the person who built that agent that's accountable." That structure collapses if companies slash headcount and hand too much to AI without sufficient human oversight.
Several summit attendees told me — with evident discomfort — that agentic tools were making their junior hires look increasingly redundant. But Priest pushed back on the impulse to stop hiring at the entry level. Despite a growing body of research showing entry-level roles in decline, he argued companies still need to fill those positions — and can't fully automate what they represent.
Also: The great AI skills disconnect — and how to fix it
"Up until now, the structure of work has been based on a pyramid — many workers, fewer leaders at the top," he said. "I see the ideal structure moving more toward an hourglass: lots of entry-level hires at the bottom who you can and should invest in, and fewer middle managers." Junior employees, he argued, tend to be more motivated, more adaptable, and more cost-effective to reskill than mid-level managers who've grown comfortable in established roles.
That said, he cautioned against letting junior employees overcorrect in the other direction — leaning on AI so heavily that they never develop the underlying judgment that makes them valuable long-term.
"With senior employees, there's maybe not enough adoption," he said. "And with junior hires, there's likely too much."