There is a particular kind of press release you learn to recognise in this line of work: the one where a firm that has spent decades billing by the hour suddenly insists it was a technology company all along. KPMG US chief executive Tim Walsh said as much at the World Economic Forum in January, telling Business Insider: “there is no doubt that our firm is a tech company that delivers now on audit, tax, and advisory services.” A decade ago, KPMG’s own vice chairman of advisory, Rob Fisher, says the firm would have described itself as “smart people doing smart things,” billed by time and materials (Business Insider). That is not a rebrand. That is a different business model wearing the old firm’s name.
I have spent enough years around data and analytics advisory work to know how much of consulting has always run on the deck. You diagnose a client’s problem, you structure the argument, you build the fifty-slide narrative, and the client’s team executes (or does not). AI is not the first thing to threaten that model, but it might be the first thing to threaten it structurally rather than cyclically. Clients increasingly want tools, agents and managed services they can run themselves, not a report telling them what to build. Fiona Czerniawska, chief executive of the sector-intelligence firm Source Global, put it plainly: “the more they’re perceived to be a technology firm, the more likely they are to win business” (Business Insider). Perception is doing a lot of work in that sentence, and the interesting question is how much of it is catching up to reality.
The hiring numbers tell a straighter story than the branding
Rhetoric is cheap. Headcount is not. Accenture has added nearly 40,000 AI and data professionals over the past two years, on top of a firm that already sold itself on technology capability (Business Insider). That ranking held up in Forbes’ 2026 survey of the World’s Best Management Consulting Firms, produced with Statista from a global panel of 5,400 consultants and 3,700 clients across 31 countries, where Accenture placed first among management consulting firms, ahead of Deloitte and McKinsey (Forbes). Accenture’s chief strategy and services officer, Manish Sharma, frames the pitch in a way I recognise from every data project that has ever stalled on bad plumbing: “technology is almost never the constraint. The constraint is what sits underneath, because layering AI on top of broken infrastructure, siloed data and outdated workflows only automates waste” (Forbes). That is the correct diagnosis. It was also the correct diagnosis for the enterprise data programmes I have seen, well before anyone called it AI.
PwC has rebuilt its training curriculum around 30 core skills, split evenly between AI-centric and human-centric competencies, and its chief people and inclusion officer for the US, Yolanda Seals-Coffield, calls both sets “extremely critical” to its success (Business Insider). In February the firm introduced an engineering career track, its first new track in 170 years of existence, which its then chief technology and innovation officer Matt Wood described as a “signpost” for where the firm expects value to be created next (Business Insider). EY has added 61,000 technologists since 2023 and is using AI internally as a training assistant to help staff anticipate how their own roles will evolve (Business Insider). Deloitte has gone further still, scrapping titles like “analyst” and “consultant” altogether for more specific US job titles (Business Insider). When a firm changes what it calls its own people, it is usually further along in changing what it sells than its marketing admits.
What the work actually looks like now
The clearest illustration in the reporting is not a partner but a working consultant. Allison Heithoff, five years into her career at the midsize firm West Monroe, describes her early years as “hands-to-keyboard work”: gathering client data, configuring systems, developing features, testing them, deploying them (Business Insider). AI now does a meaningful share of that work, and her role has shifted toward strategy and technical judgment, using AI to code and build client solutions in ways she says she never expected when she studied business administration (Business Insider). That is the pattern I would expect from the inside: the tasks that used to justify a junior headcount get automated first, and what is left concentrates around the judgment calls, which is exactly the part that is hardest to templatise and the part firms are now racing to hire and train for.
None of this means the strategy deck is dead. What is under real pressure is the economics that made the deck profitable: a pyramid of junior staff billed by the hour to produce artefacts a client’s own AI tools can increasingly draft themselves. Firms that can charge for subscription-style products, agents and multi-year implementation work have a path forward. Firms whose value proposition stops at the recommendation do not.
Where this leaves the client relationship
I keep coming back to Fisher’s line about KPMG: “we have to become more of a technology organization because clients want to consume our expertise that way” (Business Insider). That is not really a statement about AI. It is a statement about delivery format, and it happens to be true of most enterprise software relationships I have worked around, not just consulting ones. Clients used to buy a person’s time and judgment in discrete chunks. Increasingly they want an ongoing capability, embedded and running, that they can hold accountable the way they hold a piece of software accountable: uptime, output quality, measurable results, not billable hours.
That is a harder thing to sell than a project, because it requires the firm to stand behind an operating system rather than a set of recommendations, and it changes who gets promoted, since builders and integrators start to matter as much as the partners who used to win the room. It also changes what a junior hire needs to know on day one. Five years ago a graduate consultant needed a spreadsheet and a story. Now, judging by the hiring and training decisions at Accenture, EY, PwC and Deloitte, they increasingly need to be able to build something that works.
Whether that produces better client outcomes is a separate question from whether it produces a durable business model for the firms making the bet. The Big Four and MBB have survived stranger transitions before, from the rise of in-house strategy teams to the growth of boutique specialists. But those were competitive shifts within a stable business model. This one is a shift in the model itself, and the firms that get it wrong will not know it from a bad year. They will know it from a slow one, as clients quietly stop asking for decks and start asking for systems that already work.