🗞️ Opening Line
Happy Friday! The Big Four accounting firms just completed a sweep — Deloitte, PwC, and now KPMG have all deployed Claude across their global workforces. GPT-5.5 just cut hallucinations by 52%. And IBM committed $5 billion to secure the foundations the AI economy runs on. Let's close the week.
💡 The Big Update
The Big Four Just Standardized on Claude. All of Them.
Three weeks ago we covered PwC deploying Claude to 30,000 US professionals. Two weeks ago it was Deloitte — 470,000 employees. This week it's KPMG.
KPMG and Anthropic announced the KPMG Digital Gateway Powered by Claude, bringing Anthropic's frontier AI directly into KPMG's client delivery platform, with an initial focus on tax clients and private equity firms. Every one of KPMG's 276,000+ employees globally will gain access to Claude across 138 countries and territories. LLM Leaderboardtechmeme
But the headline number isn't the story. The story is how they're deploying it.
They're not just handing out login credentials and hoping people figure it out. They're embedding Claude inside KPMG Digital Gateway — the platform their people and clients use to do actual work. They're starting with Tax and Legal. They're expanding from there. They've built a governance framework around it. National Today
Corporate tasks such as configuring an AI agent to adapt to changing tax regulations — a process that previously required weeks of manual engineering and juggling multiple tools — can now be generated within minutes inside Digital Gateway. Tax Foundation
That last sentence is the one worth sitting with. Weeks to minutes. Not for a demo. For real client work at one of the world's largest professional services firms.
The pattern across the Big Four is now impossible to ignore. These are not pilots. They are not experiments. They are infrastructure decisions — the same kind that determined which software companies dominated the last twenty years. The firms that get AI embedded into how work actually gets done will have a structural advantage that compounds over time.
For business owners watching from the outside: the firms that advise the world's largest companies just made their AI bet. It's Claude. That's not a coincidence — it's a signal worth paying attention to.
⚡ Quick Commits
GPT-5.5 Is Now the Default in ChatGPT — and It Hallucinates 52% Less OpenAI moved GPT-5.5 Instant to the default model in ChatGPT, reporting 52.5% fewer hallucinations on high-stakes prompts spanning medicine, law, and finance. That reliability improvement matters more than any benchmark score. The single biggest barrier to trusting AI with real business decisions has been hallucination — getting confident wrong answers. A 52% reduction on high-stakes prompts is a meaningful step toward AI you can actually rely on. TechCrunch opentools
IBM Just Committed $5 Billion to Secure the Foundations AI Runs On IBM's $5 billion commitment reflects a growing recognition that open-source security is now critical infrastructure for the AI economy. Most AI products — including the tools your business uses — are built on open-source components. If those components are compromised, everything built on top of them is too. IBM is betting that securing that foundation layer is one of the most important infrastructure plays of the decade. They're probably right. TechStartups devFlokers
🛠️ The Tool Drop
Claude for Enterprise Given this week's news, it's worth knowing what KPMG and the Big Four actually deployed. Claude for Enterprise is Anthropic's business-grade offering — the same Claude you may already use, but with enterprise security controls, SSO, expanded context windows, admin controls, and usage analytics. For growing businesses that are starting to think seriously about AI as infrastructure rather than a productivity experiment, it's the version worth evaluating. Not a solo tool — a platform decision. claude.ai/enterprise
🎯 The So What
Deloitte. PwC. KPMG. Three of the four largest professional services firms in the world — firms that collectively advise most of the Fortune 500 — have now standardized on Claude as their enterprise AI platform.
That's not a coincidence. It's not a marketing story. It's a procurement decision made by organizations that run some of the most rigorous vendor evaluations in the world.
The question it raises for every other business is the same one it raises every week: where are you in your own AI adoption journey? The gap between organizations that have embedded AI into how they actually work and those still experimenting with it is growing. The Big Four just made their position clear.
Have a great weekend.
— The Changelog
Not subscribed yet? Get The Changelog free, three times a week. Subscribe here
Know someone who'd find this useful? Forward it along.