Tuesday, 25 August 2026
Artificial intelligence is beginning to change the role of virtual data rooms in real estate and corporate transactions, moving them beyond secure document repositories towards platforms capable of analysing the information held inside them. The development could significantly alter how investors, lawyers and advisers approach due diligence, particularly on large transactions involving thousands of documents.
Drooms has expanded its transaction platform with Drooms Intelligence, an AI-based system designed to analyse documents held across an entire data room. Rather than searching through individual files, authorised users can ask questions about the material and receive responses connected to the underlying documents. The approach addresses one of the practical problems emerging as professional teams adopt generative AI. Confidential leases, financial information, technical reports and corporate documents may otherwise need to be removed from their protected transaction environment before they can be analysed using external applications, potentially introducing additional questions surrounding confidentiality, permissions and information governance.
Drooms is instead keeping the analysis within its European platform. Existing access rights continue to determine which information users can reach, while the system is designed to show the documents supporting its responses. “If deal teams first have to copy confidential documents out of the data room – or, worse, have them transferred automatically and without control to third-party providers – in order to use AI effectively, little has been gained,” said Alexandre Grellier, CEO of Drooms. “With Drooms Intelligence, we bring the analysis to where the data is already protected – and where it should remain.”
The development is part of a wider change across transaction technology. Virtual data rooms have traditionally concentrated on secure storage, access control and the exchange of confidential information between buyers, sellers and their advisers. AI is creating the possibility of analysing that material without leaving the same controlled environment. Other transaction technology providers are moving in a similar direction, suggesting that embedded AI analysis is gradually developing from an additional feature into a new area of competition between data-room providers.
The potential application to commercial property is substantial. A portfolio acquisition can involve hundreds of leases alongside technical reports, energy certificates, financing documents and other records. Before an investment committee can evaluate the transaction, teams must extract and compare large quantities of information from those documents. Drooms’ new platform initially provides two principal functions: one enables users to question information contained throughout the data room, while another converts recurring information contained in leases and energy certificates into structured tables that can be reviewed and used for portfolio analysis.
For real estate investors, automated extraction could be particularly valuable when examining large portfolios. Rent levels, lease expiry dates, break options, indexation provisions and other information can be distributed across hundreds of separate documents. Technology capable of identifying and organising these details could substantially reduce the amount of manual work required before analysis begins.
The more important issue, however, is whether the information can be trusted. Research published by Datasite in 2026 indicates that AI is already becoming established within transaction processes, with half of surveyed dealmakers regularly using or having substantially incorporated the technology into due diligence. Accuracy remains one of the principal concerns, while human checking continues to play an important role in validating results.
This creates a different standard for transaction AI than for general productivity applications. An incorrect summary in an internal document may be inconvenient, but an incorrect interpretation of a lease break, guarantee, liability or financial figure during a major acquisition could affect pricing and investment decisions. The ability to trace an AI-generated conclusion back to the original document is therefore becoming an important part of professional applications, allowing transaction teams to use technology to locate and organise information without treating the generated response as the final authority.
Security is similarly becoming a competitive issue. The question for deal teams is no longer simply whether AI can analyse a document, but where that analysis occurs, which systems receive the information, whether existing permissions remain effective and whether confidential transaction material is used outside the controlled environment. This is particularly relevant for European transactions, where data protection, confidentiality and increasingly detailed AI governance requirements can influence how businesses introduce generative technology into professional workflows.
The development does not remove the need for professional judgement. AI may be able to identify particular clauses across hundreds of leases, but deciding whether those provisions materially change the valuation or risk profile of a portfolio remains the responsibility of lawyers, advisers and investment teams. Instead, the technology could change where professional time is spent. Large amounts of due diligence work involve finding, categorising, comparing and transferring information before specialists can analyse what it means. Automating part of that process could allow transaction teams to spend more time examining exceptions and risks rather than locating the information in the first place.
That could eventually influence the economics of transactions. Large due diligence exercises have traditionally required substantial teams to process documentation within relatively short transaction timetables. If AI can reliably undertake part of the initial extraction and comparison work, staffing requirements and advisory workflows could change. Commercial property is particularly suited to this type of automation because much of the information involved in portfolio transactions is repetitive but financially important. Hundreds of leases may contain broadly similar categories of information while differing on individual provisions that can materially affect value.
The emergence of AI-enabled data rooms therefore represents more than the addition of another technology function. It is beginning to change the purpose of the data room itself. For years, the industry’s main challenge was creating a secure environment in which buyers and advisers could access confidential transaction documents. The next stage is turning those documents into information that can be interrogated and compared without compromising the controls surrounding them.
As competing platforms introduce similar capabilities, simply offering AI is unlikely to remain a meaningful differentiator. The more important competition will be around accuracy, security, integration and the ability to demonstrate exactly where an answer came from. For real estate investors and advisers, that could make the data room considerably more important to the transaction process. Instead of being the digital filing cabinet through which due diligence passes, it is gradually becoming part of the analytical infrastructure through which investment decisions are made.
Source: CIJ.World Research & Analysis
