How Many Deals Use AI Screening Today?

Updated August 2026 · 6 min read · Deal Room Intelligence Series

What we actually checked

The percentages you encounter — some large share of dealmakers “using AI,” some small share reporting no use at all — share a set of properties that should stop you citing them:

• Vendor-sponsored fieldwork. The survey was commissioned by a party with a commercial interest in the result. This does not make the fieldwork dishonest, and the research firms involved are often reputable. It does mean the question wording, sample frame and publication decision all sit with an interested party.
• Undisclosed sponsorship at the point of citation. The figure gets repeated in articles and decks with the sponsor's name dropped, so by the third hop it reads as independent market research.
• Sub-statistics that do not appear in the source. We traced a widely-repeated headline figure back to its cited source and could not find it in the document at all. That is not a rounding discrepancy; it is a claim with no origin.
• “Using AI” is undefined. This is the fatal one. It covers a partner using a general chatbot to draft an email and a firm running every document through a structured screening pass. Aggregating those produces a number that means nothing.

What can be established with dates instead

Adoption is unmeasurable from where we sit. The governance response to adoption is fully documented, dated and checkable — and it is a better proxy, because institutions do not write rules for things nobody is doing.

That is a governance surface going from essentially nothing to four significant instruments in under three years. Every element is a dated public document you can verify in a minute, which is more than can be said for any adoption statistic in this market.

It also answers the question the adoption number is usually a proxy for — is this real, or is it hype? — with better evidence. Standards bodies and bar associations are conspicuously slow-moving. They do not produce guidance for practices that are not happening.

Why the question misleads even when answered

Suppose someone handed you a reliable figure. It would still be close to useless for a decision, for three reasons.

It aggregates incompatible things. A firm where one partner uses a chatbot and a firm running structured screening across every data room both count as adopters. The gap between those two states is the entire decision you are trying to make.

It measures purchase, not practice. Software bought is not software used. Adoption surveys ask about procurement because procurement is easy to report. The failure mode in this category is not failing to buy; it is buying and then using the tool on two deals a year because verification was too expensive or the output was not trusted.

Peer adoption is a poor guide in a market with no published benchmarks. If nobody publishes prices, throughput or error rates, “most firms are doing it” tells you that most firms made a decision under the same information vacuum you are in. That is herd data, not evidence.

The questions worth asking instead

“What share of our own data rooms do we currently examine?”

Not read — examined. Take your last three transactions, count documents in the room against documents anyone opened. Most firms have never calculated this, and it is the number that actually bears on the decision, because it quantifies the risk you are already carrying.

“How were the unread documents selected?”

If the honest answer is folder structure, upload order, or whoever flagged something in a call, then selection is currently uncorrelated with risk. That is the gap a screening stage closes, and it is measurable internally without any market data at all.

“What did our last abandoned deal cost, and when did we find out?”

The spend between kick-off and the moment the disqualifying issue surfaced is real money from your own history. It is the strongest number available for a business case, and no survey is required to produce it.

“Could we reconstruct why a document was not escalated?”

If a claim arrived tomorrow about a provision in an unread document, what does the record show? For most firms the answer is nothing, and that is a position worth changing regardless of what anyone else is doing.

What we can say about direction, honestly

Two structural observations that need no survey:

Data rooms are getting larger, not smaller. More disclosure, more contracts, more electronic records per transaction. The volume of material per deal is rising while the number of hours a senior reviewer has is fixed. That gap widens mechanically, and something has to absorb it — either coverage falls, or triage moves to a system.

The requirement to show your work is hardening. See the table above. Whatever the adoption rate is, the standard for demonstrating how a conclusion was reached has risen on a documented timeline, and that is where screening pays — not in speed, but in producing a record of what was examined.

Bottom line

Nobody can tell you what share of deals use AI screening, including us, and the confident figures in circulation are sponsored or untraceable. The dated governance record is checkable and tells you more: four significant instruments in under three years, which is not what a market does about a fad.

Decide on your own coverage gap instead. It is the only number in this discussion that is both knowable and relevant to you — and unlike the adoption rate, you can measure it this week.

Measure your own coverage gap instead →