For decades, due diligence has been shaped by one unavoidable limitation: there has never been enough time to examine everything. Buyers, investors and advisers have traditionally worked within tight timeframes, reviewing selected contracts, focusing on the largest customers and concentrating on the information considered most material because examining every document, transaction and commercial relationship was simply impractical.
Artificial intelligence is changing that reality. Deal teams can now process far larger volumes of information in much shorter timeframes, allowing acquisitions to become both faster and broader in the way they are assessed. The real value of AI is not that it makes decisions on behalf of investors, but that it allows them to see far more of a business before committing capital.
At the same time, AI has made it easier to become confidently wrong. Today, thousands of contracts can be reviewed simultaneously, with unusual clauses, missing documentation and potential risks identified automatically. Financial records can be compared against supporting information, while customer concentration, churn patterns and operational inconsistencies can be highlighted long before a transaction reaches completion. Rather than relying on representative samples, investors increasingly have the opportunity to examine the entire picture.
This wider lens matters because many of the risks that undermine acquisitions rarely appear within headline financial performance. A business may demonstrate strong revenue growth and healthy margins, yet still carry significant exposure through customer concentration, founder dependency or operational weaknesses that only become visible when information is examined in greater depth.
Due diligence is also no longer confined to the data room. Public filings, hiring activity, customer reviews, digital engagement and wider market signals can now be analysed alongside management information, allowing buyers to test whether a growth story is genuinely supported by external evidence. However, the expansion of information introduces a new challenge: false confidence.
One of the newest mistakes emerging within due diligence is the tendency to accept AI-generated summaries without sufficiently questioning the information from which those conclusions have been drawn. Reports can appear highly convincing and provide a level of certainty that may not always be justified.
Technology can identify patterns, organise information and highlight anomalies, but it cannot assume responsibility for the decisions that follow. The most important questions within any transaction frequently sit outside the data itself and require experience, judgement and commercial instinct.
Why is the business being sold at this particular moment? How dependent is performance upon a small number of individuals? Are the cultures genuinely compatible? Will customers remain after the founders leave? These questions rarely appear within financial statements, yet they often determine whether a transaction ultimately succeeds or fails.
In my own experience across property, hospitality and growth businesses, the most valuable lessons have rarely come from the numbers alone. One business appeared highly attractive from a financial perspective, yet a closer examination revealed that much of its value depended upon a small number of relationships and key individuals. The principal risk was not the profitability of the business, but whether those relationships would survive a change in ownership.
Similarly, there have been situations where an investment case depended heavily upon one particular assumption proving correct. In those circumstances, the quality of the investment decision depended not upon the sophistication of the financial model, but upon the willingness to challenge that assumption repeatedly and, if necessary, walk away from the opportunity altogether.
The importance of human judgement becomes even clearer once a transaction has completed. Many acquisitions do not fail during due diligence itself. They fail during integration, where leadership, communication and execution determine whether value is ultimately realised.
Artificial intelligence can improve visibility and accelerate analysis, but it cannot build trust between management teams, retain customers or lead people through organisational change.
Looking ahead, due diligence is likely to become increasingly continuous rather than event-based, with businesses moving away from static assessments towards ongoing monitoring of financial performance, customer activity and operational indicators. Alternative sources of data will become more important, while data rooms themselves may increasingly be designed for machine analysis as well as human review.
As the cost of analysis continues to fall, smaller transactions may become more commercially viable because buyers will be able to undertake deeper diligence without incurring the costs that have traditionally limited access to high-quality analysis. This may create opportunities for smaller investors, entrepreneurs and owner-managed businesses that previously lacked the resources available to larger institutions.
Artificial intelligence has undoubtedly made due diligence faster and deeper, but it has also made it easier to become confidently wrong. The challenge for investors is therefore not simply understanding what the technology can do, but recognising where its capabilities end and where human judgement must begin. The cost of analysing a deal is collapsing. That does not make judgement less valuable. It makes it the whole game.
