Netherlands - Market Insights

AI makes deals smarter, if you do the thinking first

Artificial intelligence (AI) is rapidly reshaping the mergers and acquisitions (M&A) landscape in the Netherlands. Advanced AI tools, from machine learning systems to generative applications such as Microsoft Copilot, are increasingly integrated into both the legal M&A transaction process and the business operations of deal targets. More than half of recent Dutch M&A deals now involve some form of digital transformation or AI component. Global adoption of generative AI has also surged: around 33% of organisations used it regularly in 2023, rising to 65% in 2024 and 69% in 2025. Nearly half of dealmakers now expect to acquire an AI-driven company in the near term, a substantial increase compared to 2023. AI has therefore become not only a driver of deal activity, but also a tool for executing transactions.

At the same time, the technology brings new legal risks and challenges. While AI can significantly enhance efficiency, it does not replace human judgment in legal M&A practice. This contribution examines how AI is influencing Dutch M&A transactions, focusing on AI‑related legal risks as well as the role of AI within the transaction process itself.

AI-related legal risks

Because AI systems rely heavily on data, issues of the legality of data use and quality have become central in due diligence. Buyers now routinely seek assurance that the target company has the rights to use the data used to train and operate its AI models, and that such data was collected and processed in compliance with applicable laws, including the GDPR. As a result, due diligence increasingly involves deep scrutiny of the target’s datasets, data governance measures, and privacy practices.

Intellectual property is also under heightened examination. This concerns not only the AI code and algorithms, but also the outputs they generate. Legal uncertainty remains around whether and to what extent AI-generated outputs qualify for IP protection, since works are required to be the maker’s own intellectual creation in order to be eligible for copyright protection in the Netherlands. Buyers must therefore be alert to potential ownership gaps, as they will expect — post-closing — to use the target’s AI systems and their outputs with the required legal protection and without legal obstacles. In modern M&A, data and AI-related IP have become both the crown jewels and the potential pitfalls.

Regulation is tightening as well. The European Union’s AI Act, adopted in 2024, introduces a risk-based regulatory framework for AI systems and bans certain high-risk AI practices. From 2025 onward, companies operating AI in the EU have begun to face new requirements under the European AI Act, starting with rules on prohibited AI practices and AI literacy, and followed by broader obligations around transparency, risk management, and human oversight. Dutch regulators, such as the Dutch Data Protection Authority (Autoriteit Persoonsgegevens), which has identified AI as a key focus area for 2026–2028, have signalled active enforcement, with a focus on data usage, algorithmic bias, and privacy. Buyers are therefore examining whether targets use AI in a legally compliant and responsible way and whether hidden regulatory risks — such as potential fines, injunctions, or required system changes — may arise. For example, if a target deploys or provides a high-risk AI system, buyers will want to know if the requirements of the AI Act (and other applicable rules and regulations) are complied with.

These risks have prompted buyers to seek enhanced contractual protection in transaction documentation. Sellers are increasingly required to warrant that the target’s use of AI complies with all applicable laws and regulations (including the European AI Act) as well as relevant ethical guidelines. In addition, buyers may require sellers to warrant that the company owns, or has valid rights to use, AI‑generated works, and that it has lawful rights to the data used in its AI systems. It is also becoming more standard to include warranties confirming that confidential company data has not been improperly entered into external AI tools, such as publicly accessible chatbots. Robust representations and warranties to this effect are now routinely included in transaction documentation. Where due diligence identifies specific AI-related risks, the parties may negotiate tailored indemnities.

Together, these contractual mechanisms ensure that AI-related risks are explicitly addressed and clearly allocated. In practice, negotiating such clauses forces buyers and sellers to think carefully about how the target uses AI and to align expectations on how potential issues will be handled post-transaction. In our experience, discussions around AI warranties and indemnities often reveal misunderstandings or technical complexities that benefit from human clarification; the kind of nuanced analysis no algorithm can replicate.

AI in legal M&A transaction processes

AI has made an irreversible entry into legal M&A practice. Its principal advantages — speed, efficiency, and scalability — mean it is now routinely used across a wide range of legal tasks, including contract drafting, legal research, due diligence review, and text optimisation. AI is transforming how transaction teams handle large volumes of information, enabling lawyers to work faster and more efficiently than ever before.

The most significant impact to date is visible in the due diligence process. Advanced AI tools such as CoCounsel, Lexis+, and Harvey can review and analyse vast numbers of documents within a fraction of the time required by human reviewers. They identify unusual clauses, flag risks, and detect inconsistencies or anomalies that might otherwise go unnoticed or require weeks of manual review. AI-driven due diligence frees deal teams to focus on genuinely critical issues rather than mechanical document review, accelerating transaction timelines and reducing client costs.

At the same time, AI has important limitations. Generative AI produces text that appears coherent and authoritative, but it does not actually understand legal context, commercial realities, or downstream consequences. It may generate output that is plausible-sounding yet factually incorrect, incomplete, or misleading, often referred to as “hallucination.” AI therefore cannot be relied upon as an autonomous decision-maker in legal M&A processes.

The quality of AI-generated output also depends heavily on the quality of the input. An AI system trained on incomplete, biased, or outdated data will produce flawed results, and poorly framed or overly broad prompts will yield unhelpful answers. The principle of “garbage in, garbage out” applies fully. Effective use of AI requires guidance from experienced professionals who understand both the legal issues at stake and the limitations of the technology. While AI can identify patterns or flag potential issues, it cannot determine whether a particular clause represents a material legal or commercial risk, nor can it weigh competing considerations or exercise judgment in the strategic context of a deal. An AI tool may spot an unusual indemnity provision, but only a lawyer can assess its significance, negotiate appropriate amendments, and advise the client on the associated risks.

In practice, this means that AI should for now be used in legal M&A practice only in a supporting role. It is particularly effective for preparing initial drafts, outlines, or summaries that are subsequently reviewed, verified, and refined by legal advisors; double-checking and enhancing human work products; and navigating large document sets more efficiently by pointing advisors to the relevant precedents, documents, or clauses. AI has, however, already proven to be a true game changer in areas such as text drafting (correcting typos and grammar errors) and layout optimisation, where it can significantly reduce the time required to finalise transaction documents, thereby generating efficiency gains for lawyers and clients alike.

Conclusion

AI has become an integral part of M&A practice in the Netherlands, both as a strategic driver of transactions and as a powerful tool within the transaction process itself. While AI offers clear efficiency gains — particularly in due diligence and document handling — it also introduces complex legal, regulatory, and contractual risks that require careful management. Issues around data ownership, intellectual property, and compliance with evolving frameworks such as the EU AI Act now form a core part of M&A due diligence and deal negotiations. As a result, enhanced warranties, representations, and indemnities addressing AI-related risks are increasingly standard. Despite its growing capabilities, AI cannot replace human legal judgment. Its outputs require critical assessment, contextual understanding, and strategic decision-making by experienced professionals. For the foreseeable future, AI’s role in M&A is therefore best understood as a sophisticated support tool; one that enhances, but does not replace, human expertise.