A report titled 'AI and the Future of Work,' released last week by the Infosys Knowledge Institute, the research arm of Infosys, found that more than two-thirds of employees use AI at least once a day. The survey covered more than 2,600 white-collar workers. Corporate users reported that AI saves them an average of four hours per week, and most said they redirect that freed-up time toward handling new tasks or producing higher-value work.
AI experience diverges even within the same company
The share of employees who said they always have the AI tools they need was over a third among senior executives, but only 19% among junior employees. Infosys noted that access to tools and training tends to be more open at the executive level, which can lead to greater confidence and a sense of security among executives—while at the same time distorting their perception of how effectively AI is actually being used by employees below them. Training methods were also flagged as an obstacle, with some workers feeling that the upskilling opportunities provided by their companies are disconnected from their day-to-day work.
Transformation goals differ by level as well
Nearly half of senior executives said their company's goal is to redesign most or all of its work and processes around AI, but fewer than a quarter of middle managers gave the same answer. The report described this as some of the starkest evidence that senior leadership has failed to communicate its vision for the future of enterprise AI, adding that execution cannot be expected if middle managers don't agree with top leadership's transformation vision. Infosys viewed middle managers—who sit at the center of both strategy and execution—as being in a position to shape how work actually gets done and to provide the feedback needed to refine adoption plans.
Seth Robinson, Vice President of Industry Research at CompTIA, said CIOs should prepare on the assumption that AI proficiency may be low not only among existing employees but also among new hires. He noted that organizations cannot assume individuals arrive with AI knowledge that's immediately applicable to the organization, making it all the more important for companies to first clearly define exactly what they want in terms of AI capability.
What Korean companies should measure in their own organizations
Organizations pursuing AX transformation can translate the two diverging metrics from this survey into numbers specific to their own company. The first is tool access. Breaking down issued AI tool accounts by job level reveals, in concrete numbers, the gap between the adoption level executives assume exists and the adoption level that practitioners actually experience. The second is training. Checking whether the hands-on exercises in last quarter's AI training were drawn from the actual work trainees perform, or were generic examples instead, can reveal whether upskilling is genuinely connected to day-to-day tasks.
The quickest thing to try is simply asking middle managers directly. This week, try asking them to answer in a single sentence how far your company intends to go in transforming work and processes with AI, and check whether the sentence that comes back matches the one executives have in mind.