Key takeaways

• “Skilling now should happen at the speed of innovation,” or organizations risk leaving workers behind
• AI transformation often stalls because people lack the skills and confidence to use the tools effectively
• Leaders’ readiness matters: “The whole success comes from leadership training”
• Without intentional design, AI could create “a two-tiered workforce; an AI enabled and an AI not enabled [one]”
• Pathways may need more structure: competency baselines, coaching and time-bound learning tied to outcomes
• Access is also about aspiration: educators need real exposure to jobs and skills to guide learners credibly
• Scaling requires breaking silos between employers, educators and governments, and building career navigation systems.

“It’s not a question of whether AI is actually reshaping work or reshaping education,” said moderator Neil Allison, Head of Education, Skills and Learning Mission at the World Economic Forum, said. “The shift is happening, whether we like it or not.” The question now is “the speed, the scale and who benefits individuals, countries and the organizations.”

With Srimathi Shivashankar, Corporate Vice President and Business Head, EdTech Services at HCLTech, Vishaal Gupta, President, Enterprise Learning and Skills at Pearson and Andrew Baird, CEO at Education For Employment, the discussion moved quickly from technology to capability: the human systems that determine whether AI becomes advantage, anxiety or inequality.

The missing link isn’t access to AI. It’s people’s ability to use it

Gupta pointed to a widening gap between the noise around AI’s promise and the reality inside organizations. He highlighted that most AI transformation projects are failing, but to take advantage of the technology requires “humans, employees, partners and customers” to have the necessary skills.

That gap is becoming more visible as leaders confront a flood of tools. “They’re getting overwhelmed by technology,” said Gupta. “Every day, there new language models are getting released,” but the question remains of how to use them effectively.

For him, the goal is straightforward: AI needs to deliver the same business outcomes leaders are already measured on. “We are after growth. We are after better employee experience. We are after better customer experience,” he said. “AI is no different…It needs to deliver precisely the same KPIs.”

“Skilling now should happen at the speed of innovation.”

Shivashankar framed the workforce challenge as a timing problem as much as a content problem. “Skilling should happen at the speed of innovation,” she said, arguing that if learning lags innovation, “you’re going to leave a lot of people behind.”

She described three shifts she is seeing inside organizations:

• Jobs are being rewritten with technology embedded inside them
• Skilling is moving from generic training to job-competency alignment
• Co-creation is replacing solo redesign

The operational implication wasn’t “more content.” It was a different development model: shared resources, shared design and learning that keeps pace with real work.

The risk is a two-tier workforce: “AI enabled and an AI not enabled.”

When Allison asked whether AI might deepen inequality even as it expands opportunity, Baird said: “Yes. I’m concerned that we end up with kind of a two-tiered workforce going forward, an AI enabled and an AI not enabled.” Avoiding that outcome will require “a lot of intentionality.”

He brought in a telling data point from his organization’s base of employers: “We recently surveyed [some of our] employer partners…Less than half of them cited AI skills as important to their business.” That disconnect matters, he argued, because if employers don’t value AI skills, they won’t hire for them and jobseekers won’t prioritize them. “They’re not recognizing…what the benefits could be,” he said. “So, they’re not looking for that in their new talent.”

At the same time, Baird pointed to a countervailing opportunity: “Very often the youth who are entry level…become the AI experts at their businesses…they immediately take that position.” The entry point can still be a leap, but only if organizations build pathways that don’t assume prior access or social capital.

Executive anxiety is real and it may be slowing adoption

The panel repeatedly returned to a human dynamic that rarely makes it into AI strategy decks: fear.

Gupta referenced a recent survey of senior executives: “They asked them the emotional feeling they have about AI…[and the majority of] those people said they’re nervous about it.”

Mirroring Baird, he also underscored a generational contrast that has become common in enterprises: “The young people [are] getting much faster…But I think the execs who run the companies today…are struggling with...AI.”

Shivashankar added a related angle: the learning challenge is not only in acquiring new skills but letting go of old ones. “The most difficult thing they have today is how to unlearn to learn,” she said. For experienced professionals, “they’re so used to doing certain things the same thing…suddenly…they have to just shut off a lot of things and get into a new world.”

Leaders can’t outsource readiness

When Allison asked how leaders should rethink roles, pathways and organizational design, Gupta’s answer centered on leadership capability before workforce capability.

“The whole success of AI…comes from only one thing, and that’s leadership training,” he said. He described “unlearning for leaders” as “humbling,” particularly for those who have been operating successfully for decades.

He also pushed back on a common assumption: that senior leaders already understand what AI can do. “We should not assume whatsoever that some of our business leaders… know what the power of AI is,” he said. Leadership participation in skilling programs matters, in his view, not as symbolism but as operational necessity.

Gupta also argued that the skills conversation is shifting away from purely technical credentials toward “human skills” and what he called “learning agility, learning to learn.” The premise: “If you don’t have learning muscle, you are going to struggle in the future.”

Standardization, time bounds and a more directive approach to pathways

Shivashankar described a practical constraint many large employers now face: “Many organizations…have five generations working today in the workforce.” That makes a single, fully personalized learning experience unrealistic. “We just cannot curate any of the pathways saying that I need to please all the five generations. It’s not going to happen.”

Her proposed counterweight was clearer competency baselines and more explicit pathways; more structured progressions with ROI visibility.

“I think we have no time for that,” she said, referring to purely self-directed learning at scale. Instead, she argued for “competency baseline” work, stronger “counselling and coaching,” and learning framed as a “time bound exercise” where learners “need to show...improvement.”

She also insisted this cannot be delegated to a single department. Every leader in the organization has a responsibility and needs to take accountability to help their team succeed.

Gupta reinforced the underlying tension of voluntary versus mandatory skilling with an example: when a sales organization wasn’t attending a program, the Chief Sales Officer reframed it from optional to required. “The next time everybody showed up,” he said, once people believed it could affect “performance appraisals.”

Allison offered a public-health analogy for why the “required” framing is emerging: booster shots. “If you don’t get the skilling you need, there will be consequences,” he said, “for you…[and] for your organization.”

Baird added another motivator he sees working: “The fear of missing out incentive is a really big one.”

Access isn’t only about opportunity. It’s about aspiration and guidance

Baird described access through the lens of social capital: his organization works with youth who “lack that social capital” and “access to labor market in meaningful ways.” One approach he highlighted was challenge-based engagement with employers: “Groups of four or five youth working on a challenge put out by someone in the private sector.” The goal is simple: learning that creates relationships, not just credentials.

Gupta widened the access lens further, arguing that the workforce conversation can become an “over obsession with white collar jobs.” He pointed to unmet demand in vocational sectors, such as nursing and construction, and gave a striking example from Saudi Arabia: a construction program with “200 seats” received “90,000 applications.”

The jobs, he argued, are not immune to AI, but they are more likely to be “enabled” than replaced. “These are all going to be jobs which are AI enabled,” he said. “They will never be taken over by AI.”

Shivashankar connected access to what learners believe is possible. “When you talk about access, I think we’ll also talk about aspirations,” she said. She argued that educators’ exposure to real workplaces shapes how students understand roles and pathways, describing a practice of giving “academicians also internships” so they can “go back and tell students about how to access this job and trace their aspirations…at the right levels.”

Scale requires ecosystem design, not isolated programs

In closing, the panellists converged on collaboration as the practical constraint.

Baird called for “breaking down the silos between the educators and the employers and others in the ecosystem” and building “a platform for dialog and continual, rapid change” while also pushing “lifelong learning” from slogan to reality.

Gupta argued that learning must become “more inspiring,” and more “personalized,” delivered “through…the flow of work.” He also warned that without that redesign, “we are running the risk of actually leaving a lot of people behind.”

Shivashankar put a sharper edge on the purpose: “Learning and development is seen for business competitiveness. I think it should be seen for business sustenance.” In other words: not a support function, but a core system of enterprise resilience.

Allison summarized his own system priority as “career navigation systems” that can help individuals see pathways, help employers find skills and help countries improve access to opportunity.

Workforce readiness: Determining the future of enterprise success

The panel’s most consistent message wasn’t that AI will change work. Instead, it was that workforce readiness will determine whether those changes produce growth, anxiety or exclusion. The overall message was clear: organizations are overwhelmed, leaders are nervous, pathways are unclear and learning systems are often too generic, too optional and too disconnected from real work.

The answers offered were equally concrete: build competency baselines, co-design with partners, invest in leadership learning, treat “learning agility” as a core skill and redesign learning so it is personalized and embedded in the flow of work.

Above all, the panel returned to a simple risk: without deliberate effort, AI’s benefits will cluster. “An AI enabled and an AI not enabled” workforce is not an abstract future scenario; it is the default outcome unless institutions choose otherwise.

FAQs

Is the core barrier to workforce AI readiness technology access?
Not anymore. Skills, confidence, job redesign and leadership readiness, especially as employees feel overwhelmed by rapid tool changes, are crucial.

How do organizations avoid becoming “two tiered” on AI?
By being intentional: embedding AI skills into job definitions, creating equitable pathways and ensuring learners without social capital can access practical opportunities and employer networks.

Should skilling be voluntary or mandatory?
The panel suggested a balance. Some learners won’t engage without clear incentives and expectations, but programs still need relevance, personalization and alignment to real work.

How important is learning agility?
The pace of change makes static skills insufficient. Learning to learn becomes a differentiator. Top performers today may not automatically adapt to tomorrow’s requirements.

Are vocational roles part of the AI workforce agenda?
Yes. Many vocational roles will be AI-enabled and in high demand, requiring modernized training and pathways.