Summary: The WEF’s January 22, 2026 workforce blueprint argues that AI returns require redesigning skills, roles, governance and internal mobility, not just buying models. It cites OECD estimates of up to 1.1 billion radically transformed jobs, and 86% of surveyed employers expect AI and information processing technologies to transform their business by 2030. HCLTech’s 116,000 GenAI trainees and Cynergy Bank’s measured complaint, productivity and experience gains show how training can be tied to business outcomes.
On January 22, 2026, as the World Economic Forum’s Annual Meeting was underway in Davos, the WEF published an article titled “Invest in the workforce for the AI age,” written by HCLTech’s CEO and Managing Director, C. Vijayakumar. His argument compresses into a single sentence: buying a model does not buy the returns — those arrive only when a company reworks its workforce, operating model, and governance structures at the same time.
Behind the argument sit hard numbers, all drawn from the WEF’s own research foundation: on the OECD’s estimate, technology will radically transform as many as 1.1 billion jobs within a decade, while the Reskilling Revolution initiative has set itself a 2030 deadline to bring better education, skills, and economic opportunities to 1 billion people. Place the Future of Jobs Report 2025 survey alongside those figures — 86% of employers expect AI and information processing technologies to transform their business by 2030 — and “whether to transform” stops being the question; the question is what belongs on the execution checklist.
How Big the Wave Is: 1.1 Billion Jobs, 86% of Employers
Drawing on responses from more than 1,000 large employers, the Future of Jobs Report 2025 sketches a precise picture: heading into 2030, roughly 170 million new roles are projected to appear against 92 million that disappear — a net addition of 78 million. Over the same span, 59 in every 100 workers will require reskilling or upskilling, and 63% of employers already point to the skills gap as the single biggest obstacle in the way.
Read another way, what limits progress is not what models can do but how quickly organizations can digest change. Hence the blueprint’s focus on the enterprise tier: policy and technology are in place, and it is the middle layer that refuses to move.
Three Moves HCLTech Actually Made
To show the method at work, the author reaches for his own company’s scoreboard. In the past year, HCLTech put roughly 80% of its employees through core-skills training, enrolled more than 115,000 people in digital capability building, and delivered generative AI training to 116,000 — 38,000 of them in the latest quarter by itself — while over 600 completed responsible AI training; among all OpenAI partners, the company holds the largest count of OpenAI-badged experts.
Three structural moves hold these figures up:
- A skills backbone: a taxonomy of skills anchored to value pools, ensuring every dollar spent on training corresponds to business output you can measure
- Role redesign tied to learning: repeatable, data-intensive tasks shift to AI and judgment stays with people, while guardrails — quality thresholds, bias checks, escalation paths — are built into the workflow itself rather than reviewed once results are in
- Internal mobility: talent marketplaces plus project-based staffing, so people who have just picked up new skills are genuinely redeployed into the roles that call for them
Cynergy Bank: Results You Can Measure
Nowhere in the blueprint is the ground firmer than at Cynergy Bank. Working alongside HCLTech, the European bank moved its contact center and back-office operations onto digital footing — case management, voice analytics, and generative AI agent assistance — and closed with three results: complaints fell by more than 50%, productivity rose 8%, and customer experience scores gained 25%.
The significance lies less in how large the figures are than in what kind of figures they are: every AI deployment here sat on repeatable processes with clearly defined baselines, which is exactly why the effects could be measured. That is precisely the missing piece in most enterprise AI programs — what they lack is not a bigger model but a production line whose output can be measured.
Responsible Deployment: People Keep the Final Say Where It Counts
On governance the blueprint is just as concrete: transparency throughout; human oversight kept in place wherever decisions carry high stakes — hiring, credit, safety-critical situations; and inclusive design joined with reskilling to cushion the shock of jobs disappearing. Progress should follow the rhythm of a handful of “AI missions”: each one anchored to a business outcome, initiated by a senior business leader, held jointly accountable by technology, risk, and people leadership, and released on a steady shipping schedule. Behind this line of work stands the WEF’s Centre for AI Excellence.
What Product and Engineering Leaders Should Carry Away
Three closing points. One: 2026 is when AI ROI leaves the slide deck and lands in the income statement — the same call we made in our opening-of-year outlook — and budgets will follow only the projects able to prove a productivity gain; everything else gets cut. Two: guardrails are changing address, moving out of the model layer and into the workflow layer, where quality thresholds and escalation paths turn into shipped product features rather than compliance paperwork. Three: skills data now counts as infrastructure — strip out the taxonomy and the internal mobility mechanics, and no amount of training budget ends up as anything beyond green ticks on a completion dashboard.