KPMG just went all in on AI. The Big Four firm announced a global alliance with Anthropic that puts Claude into the hands of every single one of its 276,000+ employees across 138 countries. Not a pilot. Not a sandbox. A firm-wide deployment baked into the platform where KPMG’s people actually do client work.
If you run a service business of any size, this is worth paying attention to. When one of the world’s largest professional services firms decides AI agents are ready for tax filings, legal work, and private equity deals, the bar for “we’re still evaluating” just got a lot higher.
What happened
- KPMG and Anthropic formed a global strategic alliance. Claude is being embedded inside Digital Gateway, KPMG’s main client work platform built on Microsoft Azure.
- All 276,000+ KPMG employees worldwide get access to Claude. This builds on two years of internal adoption at KPMG US, including its AI and Data Labs.
- New AI tools are launching for tax and legal clients first. Claude Cowork and Managed Agents are integrated directly into Digital Gateway so KPMG professionals can build AI capabilities without jumping between tools.
- Anthropic named KPMG a preferred partner for private equity. KPMG will deploy Claude and Anthropic’s agents into PE portfolio companies. They also launched KPMG Blaze, which embeds Claude Code to modernize legacy IT systems.
- Cybersecurity is a focus area. KPMG and Anthropic teams will use Claude to find and fix vulnerabilities in critical systems, guided by KPMG’s Trusted AI framework.
- Joint research with UT Austin is studying how humans should work alongside AI, not just whether the tech works.
The numbers
- 276,000+ employees getting Claude access globally
- 138 countries and territories where KPMG operates
- 2 years of prior Claude adoption inside KPMG US before this expansion
- Building an AI agent for tax regulation changes went from weeks to minutes, according to KPMG’s Vice Chair of Tax
5 things every service business operator should take from this
- The “wait and see” window is closing. KPMG isn’t experimenting. They’re embedding AI agents into the platform where real client work happens. If a firm that handles audits and tax filings (where mistakes have legal consequences) is comfortable deploying AI at this scale, smaller firms have fewer excuses to delay.
- The play is platform integration, not chatbots. KPMG didn’t just hand people a chat window. They embedded Claude Cowork and Managed Agents inside their existing work platform. That’s the pattern that actually changes how people work. Bolting a chatbot onto your website is not the same thing.
- Private equity is accelerating AI adoption from the top down. PE firms want their portfolio companies running leaner with AI. KPMG is positioning itself as the consultant that makes that happen. If you serve PE-backed companies, expect your clients to start asking about AI capabilities very soon.
- “Human in the loop” needs to be specific. The joint research with UT Austin makes a point worth repeating: saying “we keep a human in the loop” isn’t enough. You need to define exactly what that human is doing. Reviewing outputs? Shaping workflows? Making final decisions? Get specific or the phrase is meaningless.
- Cybersecurity is a use case, not just a concern. Most firms worry about AI security risks. KPMG is using AI to actively find and fix security vulnerabilities. That’s a mindset shift worth adopting.
The hot take
This deal (and the PwC deal Anthropic announced around the same time) tells you something important: the Big Four have picked their AI horse, and it’s Claude. That’s not a coincidence. Anthropic has positioned itself as the “responsible AI” company, and for firms where regulatory risk, client trust, and audit trails are literally the product, that positioning matters more than raw benchmark scores. The real story isn’t that KPMG adopted AI. It’s that they chose the AI company whose entire brand is built around safety and governance. Every professional services firm that competes with KPMG now has to answer the question: what’s our AI strategy? Because “we don’t have one yet” just became a competitive liability.
The Agency OS play
You don’t need to be KPMG to do what KPMG just did. The pattern is clear: pick your core work platform, embed AI agents into it, and give your team structured access with guardrails. This week, audit where your team actually spends time. Is it a project management tool? A CRM? A document management system? That’s where AI agents should live. Not in a separate tab.
Start small but start inside your workflow. If you run a law firm, look at how Claude or similar models can draft first passes of regulatory research directly in your document system. If you’re in financial advisory, build an agent that pulls relevant client data before meetings so your team walks in prepared. If you serve PE-backed companies, this is your cue to get an AI capability story ready, because your investors are about to ask for one.
The biggest takeaway from the KPMG deal is the “weeks to minutes” quote about building tax regulation agents. That compression is available to you right now. Pick one repetitive, high-volume task your team does every week. Build an agent for it. Measure the time saved. Then do it again. The firms that treat AI deployment as an ongoing practice (not a one-time project) are the ones that will look like KPMG in two years, not the ones still running pilots.