What Happens When AI Runs Due Diligence Faster Than Your Team
I’ve sat in enough deal rooms to know the rhythm. The stacks of lever arch files are gone now — replaced by data rooms and Zoom calls — but the process hasn’t fundamentally changed in twenty years. A team of expensive people spending weeks combing through documents, building models, flagging risks, and writing reports that land on someone’s desk two days before completion.
That’s about to break.
The 200-Hour Problem
A typical mid-market PE deal — say £20m to £100m enterprise value — burns through 200 to 400 hours of financial due diligence. That’s a team of three or four people working flat out for four to six weeks, reading contracts, reconciling management accounts to statutory filings, stress-testing working capital, and trying to figure out whether that spike in Q3 revenue was real growth or a one-off.
Most of that work is pattern recognition. It’s reading documents, spotting anomalies, cross-referencing data points, and asking “does this make sense?” That’s exactly what large language models are getting frighteningly good at.
What’s Already Happening
This isn’t theoretical. Right now, AI tools can:
- Parse entire data rooms in minutes. Upload 500 documents — contracts, leases, board minutes, management accounts — and get a structured summary with flagged risks in under an hour. What used to take a junior analyst a week.
- Cross-reference financial data against filings. Does the revenue in the management accounts match the CT600? Do the payroll numbers reconcile to the P11D submissions? AI doesn’t get tired at 11pm and miss the discrepancy on page 47.
- Benchmark against sector norms. Is this company’s gross margin plausible for a £30m professional services business? AI can check that against thousands of comparable transactions instantly.
- Draft the red flag report. Not the final, polished version — but a first pass that highlights the twenty things worth investigating further. That alone saves days.
The FT reported earlier this year that several major PE houses are already deploying AI in their deal pipelines. Not as a gimmick — as a competitive advantage. The firms that can evaluate opportunities faster get to term sheets first.
Where It Falls Down (For Now)
Let’s be honest about the limitations, because overpromising is how technology gets a bad name in boardrooms.
AI can’t yet reliably:
- Exercise judgment on commercial viability. It can tell you the numbers don’t add up. It can’t tell you whether the management team is good enough to fix them.
- Navigate ambiguity in contracts. A badly drafted SPA with contradictory clauses still needs a lawyer who understands what both parties probably meant — and what a judge would decide.
- Read the room. Half of due diligence is the conversation around the numbers. Why did the FD hesitate when you asked about that supplier concentration? Why is the seller pushing for a shorter exclusivity period? AI doesn’t pick up on that.
The deal room of the future isn’t empty. But it has fewer people in it, and the people who are there are doing fundamentally different work.
What This Means for CFOs
If you’re a CFO — particularly one working in PE-backed businesses or preparing a company for sale — this shift matters now, not in five years.
On the buy side: Your due diligence teams will shrink. The value moves from “can you review these documents?” to “can you interpret what the AI found and make a call?” If you’re still staffing deals the old way, you’re overpaying for hours and underpaying for insight.
On the sell side: Assume the buyer’s AI has already found every inconsistency in your data room before the first management presentation. The days of burying a problem in a footnote on page 312 and hoping nobody reads it are over. Get your house in order before you go to market, because the machines will find everything.
As an interim or portfolio CFO: This is an opportunity. The CFOs who understand how to work alongside AI tools — who can set them up, interpret their output, and add the human judgment layer on top — will be worth significantly more than those who can only do the work manually. It’s the difference between being the person who drives the car and the person who still insists on walking.
The Uncomfortable Truth
The deal advisory firms know this is coming. The Big Four are investing heavily in AI-powered due diligence platforms. The mid-tier firms are scrambling to keep up. And the boutique practices that compete on having smart people work long hours? They’re the most exposed.
None of them will tell you this publicly, because turkeys don’t vote for Christmas. But the economic logic is unavoidable: if AI can do 60% of the work in 5% of the time, the fee model has to change. And when the fee model changes, the staffing model changes with it.
The deal room in the picture at the top of this article — all glass and holograms and floating data — isn’t science fiction anymore. It’s about eighteen months away from being someone’s actual office.
The question isn’t whether AI will transform deal-making. It’s whether you’ll be the one using it, or the one it replaces.