In practice: The e& AI Graduate Programme6 is a strategic initiative designed to cultivate the next generation of digital leaders by immersing them in advanced AI and emerging technologies. The programme’s impact is evident in its ability to accelerate professional development, with participants achieving career milestones in just three years – a timeline that previously averaged seven. A core focus of the curriculum is the emphasis on practical application, equipping graduates with a sophisticated toolkit to navigate complex business challenges and fostering a mindset where AI is leveraged as a catalyst for groundbreaking innovation. This approach nurtures critical thinking and empowers graduates to deliver differentiated, high-impact outcomes, thereby driving substantial value and transformative change across the organization. AI is not affecting all workers equally. Instead, it is reshaping the various tiers of the job hierarchy in uneven and sometimes counterintuitive ways. At the entry level, many routine tasks – reporting, ticket triage and invoice matching, for example – are being automated. This creates a challenge for traditional “learning by doing”, as foundational work that once gave junior employees experience is disappearing. Many executives highlight the opportunity to speed up the process of making entry-level workers “client-ready” by shifting the focus from teaching “what and how” to teaching “why”, and from rote learning to developing critical thinking for real-world problem evaluation and resolution. Instead of spending years in the background, new hires can connect with clients earlier in their careers, with managers providing oversight and coaching.2.3 Reshaping job tiers Canaries in the coal mine BOX 5 A new paper, Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence is one of the first to quantify the impact of AI in the workplace at scale.7 Using granular, high-frequency administrative data from ADP – one of the largest payroll platforms in the US – a team from Stanford’s Digital Economy Lab observed the occupations, tasks and employment of millions of workers at thousands of private companies. They found that in occupations most exposed to genAI, employment for early-career workers – specifically those aged 22–25 – fell by about 13% relative to comparable workers, even after controlling for firm-level shocks. By contrast, employment among more experienced workers in those same occupations remained stable or even continued to grow. They also found that in less AI-exposed occupations, overall employment continued to hold up. Most interestingly, they found that AI which “augments” work is associated with stronger employment growth, whereas AI that merely automates work was correlated with a negative effect on early career employment in AI-exposed occupations. In sum: the impact is not evenly distributed. It is concentrated, it is already visible in the data and it is landing first on the newest entrants to the professional labour force. The paper was called Canaries in the Coal Mine because historically the earliest, highest-resolution signals of technological disruption appear not in national averages and not across all workers, but in specific pockets of the labour market. These pockets act like canaries in a coal mine: they react first, and they indicate where the pressure will go next if nothing is done. In previous eras, automation first transformed routine manufacturing and clerical work. Today, genAI is transforming parts of professional services, customer interaction roles, legal support, marketing, sales operations and software-adjacent analytic work – especially the entry-level segments of those jobs. The mid-career tier may face the most unexpected pressures (see Box 6). Several executives suggested that if entry-level staff can ramp up faster with AI, and if specialists can focus directly on higher-order tasks, then the coordination and supervision functions that once sustained middle managers may erode. While no company framed this as an imminent wave of job losses, the implication is that AI could hollow out parts of the middle – a possibility that runs against the popular narrative that junior staff are most at risk. This theme, while not yet supported by data at scale, was a recurring undercurrent in company responses. AI at Work: From Productivity Hacks to Organizational Transformation 13

AI at Work from Productivity Hacks to Organizational Transformation 2026