The fear that artificial intelligence (AI) will hollow out the workforce has dominated corporate boardrooms and business headlines for years, and gathered pace in early 2026. But one of banking’s most influential talent executives says the math actually points in the opposite direction — and she has the numbers to prove it.
Tanuj Kapilashrami, chief strategy and talent officer at Standard Chartered, told McKinsey in a recent interview that the global bank calculated roughly $49,000 in savings per employee reskilled and redeployed internally, compared with sourcing the same skill set through external hiring. Multiply that figure across the hundreds of roles the bank forecast would be transformed by automation and emerging technologies, and the cumulative figure, in Kapilashrami’s words, “was a staggering number.”
When reached for comment by Fortune, Standard Chartered pointed to an increase in internal hiring from roughly 30% in 2023 to over 50% by mid-2025, helping it save over $55 million in hiring costs and fees. A representative said it showed the bank is on a “good trajectory” in this regard.
The finding wasn’t a feel-good HR initiative, Kapilashrami explained. It came from a hard-nosed strategic workforce plan that Standard Chartered launched roughly five years ago, built around a deceptively simple reframe: What if skills — not job titles — became the currency of work?
“If you start thinking of skills, rather than jobs, as the currency of work, what choices would you make in how work gets done?” Kapilashrami said. The bank mapped what it called “sunset” and “sunrise” skills — capabilities that would disappear from banking within five years and new ones needed to execute the bank’s strategy — and overlaid them against existing headcount. The result was a granular dollar-value case that Kapilashrami brought directly to Standard Chartered’s board.
That board presentation shifted the conversation from how many jobs will AI eliminate to which skills they would need to build, buy, or borrow. Rather than defaulting to layoffs when automation displaced a function, the bank began identifying internal employees whose existing skill profiles could be redirected. Reskilling and redeployment, the data showed, were not only the humane choices; they were the cheaper ones.
The nuance behind the big savings
To operationalize the idea, Standard Chartered launched an internal talent marketplace approximately four years ago. Any employee can post a project online with the specific skills required; any employee across the globe can offer their expertise to fill it. As of October 2025, some 60% of employees were active on it, the bank previously told The Wall Street Journal.
In one notable example, Kapilashrami offered, the bank’s retail business in India used the platform to staff a project making its services accessible to deaf customers — drawing contributors from New York, London, and Singapore — and became one of the first Indian banks to offer deaf-friendly video banking in Indian Sign Language.
Kapilashrami was quick to clarify that her argument isn’t that AI poses no disruption. It’s that the disruption is being misdiagnosed. She said it was her firm belief that “humans will not lose jobs to machines,” but rather, “humans will lose jobs to other humans who use the machines.” That reframe places the burden on leadership — not technology — to drive transformation, and Kapilashrami argued that companies that fail to build AI fluency at every level will face a talent exodus as the gap widens between how employees experience technology as consumers versus how they experience it at work.
The implication is that the AI era is less a labor apocalypse than a skills arbitrage problem — one that companies can solve if they’re willing to invest in their existing people. Still, it’s one data point from one of the world’s most sophisticated global banks, operating with enormous HR infrastructure, a proprietary internal talent marketplace, and a chief strategy officer who quite literally wrote the book on skills-based organizations. (Kapilashrami co-authored The Skills-Powered Organization, published by MIT Press in 2024.) The conditions that apparently make reskilling cheaper than hiring at Standard Chartered may not be replicable at scale across industries.
There is also a selection problem buried inside the optimism. Reskilling works best for workers who are already closest to the skills they need to acquire — employees with strong digital literacy, educational attainment, and the cognitive flexibility to pivot into adjacent roles. The talent marketplace model that Kapilashrami describes, in which employees self-select into gigs and signal their hidden competencies, inherently favors the workers who are already most advantaged.
What the data says
The macroeconomic data doesn’t offer much comfort either. Research from the McKinsey Global Institute has projected that generative AI could automate tasks accounting for up to 30% of hours worked across the U.S. economy by 2030. Oxford economists Carl Benedikt Frey and Michael Osborne, in their landmark 2013 study of 702 occupations, found that automation disproportionately threatens middle-skill, routine-task workers — precisely the segment least likely to benefit from an internal gig marketplace History also offers a cautionary note: the promise of reskilling was loudly made during the offshoring wave of the 1990s and 2000s, and the retraining programs that followed were, by most economic assessments, deeply inadequate.
Even within the reskilling optimists’ own framework, the math raises questions. If saving $49,000 per reskilled employee is such an obvious win, why did it take a board presentation to make the case? The answer is that most companies don’t have the data infrastructure, the talent visibility, or the organizational patience to execute what Standard Chartered describes. For companies facing immediate cost pressure from AI adoption, the faster path will almost always be to reduce headcount. But despite all these noteworthy causes for concern, this example offers something somewhat rare in the corporate discourse: hope.
For this story, Fortune journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing.
