How Deloitte, KPMG, EY, and PwC Are Actually Using Generative AI in 2026
Generative AI stopped being a slide-deck buzzword at the Big 4 a while ago. Deloitte, KPMG, EY, and PwC have collectively poured more than $10 billion into AI since 2023, and by 2026 that money shows up in daily workflows, not just pilot programs in an innovation lab somewhere.
The firms aren't asking whether to use AI anymore — they're asking how fast they can get every employee using it well. If you're trying to land a role at any of the four firms, the tools your interviewer used on a Tuesday afternoon have probably changed more in the last eighteen months than in the previous decade combined.
Here's what's actually rolled out at each firm, why it matters for your applications, and which skills are worth building before your next interview.
The AI Rollout, Firm by Firm
Each firm picked a different way to get generative AI into employees' hands, and most of that spend is now going into what's called agentic AI — systems that carry out multi-step tasks with less hand-holding, rather than chatbots that just answer questions. Worth knowing before you namedrop a tool in an interview and get asked a follow-up you can't answer.
Deloitte — rolled out an internal generative AI platform called PairD back in 2024, then added Zora AI (built with Nvidia) to pitch clients on "digital workers" that can complete tasks with minimal supervision. In October, Deloitte struck a deal with Anthropic to bring Claude to roughly 470,000 employees firm-wide.
KPMG — made AI use effectively mandatory across staff, framing it as a skills problem as much as a technology one. Together with Microsoft, KPMG launched Workbench, which already connects around 50 AI agents and chatbots, with close to 1,000 more reportedly in development.
EY — built its EY.ai platform around roughly $1.4 billion in investment. In 2026, EY pushed about 150 agents into its 80,000-person tax practice for tasks like data collection and document review, with a stated target of scaling to 100,000 agents by 2028.
PwC — gave employees its own in-house assistant, ChatPwC, back in 2023 and now counts close to 200,000 regular users inside the firm, on top of more than $1 billion invested in generative AI capabilities.
Announced AI investment since 2023, by firm (US$ billions). Deloitte and KPMG figures include multi-year commitments running through 2028–2030, so treat this as directional rather than a strict like-for-like comparison.
What This Actually Changes for Your Job Search
None of this means the interview process turns into a coding test. It means the bar for what counts as "AI-literate" just moved, and interviewers can usually tell the difference between someone who's used a chatbot twice out of curiosity and someone who understands how these tools actually fit into client work.
We already broke down the specific skills firms are screening for in How AI Is Reshaping Big 4 Careers: Skills You Need in 2026 — worth a read before you update your CV. The short version: Analyst and junior consultant roles are still the main entry point, and they increasingly involve tech implementation, data analysis, and working alongside — not instead of — AI tools.
Employers want candidates who can tell which datasets are actually relevant to a question and who can sanity-check an AI-generated number before it goes in front of a client.
A cloud certification (AWS, Azure, or Google Cloud) is becoming one of the fastest ways to signal technical credibility at entry level, and all four firms now run formal partnerships with at least one of the major cloud providers. That's a more efficient use of prep time right now than trying to become a prompt-engineering expert.
Is AI Actually Replacing Consulting Jobs?
This is the question behind most of the anxiety, and it deserves a straight answer: not yet, and not at the entry level, but the shape of hiring is shifting. McKinsey has estimated generative AI can lift knowledge-work productivity by around 40%, which is exactly the kind of number that makes finance departments ask fewer people to do the same volume of work.
At the same time, PwC alone still hired around 1,300 graduates in the UK and 3,200 in the US last year — the firms are still hiring, just more selectively, and increasingly for people who can work with AI output rather than compete against it.
Nobody's job disappears because a firm bought AI licenses. The work worth paying a junior analyst to do gets narrower and broader at the same time — narrower on the routine tasks AI now handles, broader on judgment calls that still need a person.
If that question is genuinely keeping you up at night, we went deeper on it here: Will AI Replace Consultants? What Deloitte, KPMG, EY & PwC Are Really Doing.
Skills Worth Building Before Your Next Interview
Prompt literacy for real work, not novelty. Being able to draft a decent LinkedIn post with ChatGPT isn't the same skill as using an internal tool to summarize a 40-page client document accurately. Practice on real, messy inputs.
A working cloud certification. AWS Cloud Practitioner, Azure Fundamentals, or the equivalent Google Cloud badge — any of the three signals you can hold a technical conversation, which matters more than which one you pick.
Basic SQL plus one BI tool. Tableau or Power BI, paired with enough SQL to pull and shape your own data, is close to table stakes for analyst-level roles now.
Judgment over output. The most useful interview answer isn't "I used AI to do X" — it's "I used AI to do X, then checked it against Y because AI tools get this specific type of thing wrong."
Where This Leaves You
The Big 4 aren't quietly automating away entry-level jobs — they're redefining what "entry-level" means. The analysts who do well over the next few years won't be the ones who avoided AI tools out of principle, and they won't be the ones who leaned on them uncritically either. They'll be the ones who can move fluidly between the two.
That's a genuinely learnable skill, and it's one you can start building before you ever set foot in an assessment centre.
Curious how this plays out in a specific practice area? We mapped out what a data science career at the Big 4 actually looks like in 2026 — a good next read if analytics ends up being more your lane than generalist consulting.
