Artificial intelligence (AI) is no longer a niche consideration in corporate transactions. Whether as a driver of acquisition strategy, a source of unique due diligence challenges, or a powerful tool to enhance deal execution, AI is reshaping how mergers and acquisitions (M&A) transactions are conceived, valued and closed. In this advisory, Katten explores what AI means for buyers, sellers and their advisers in today’s deal environment.
AI as a Deal Driver and a Deal Risk
AI is increasingly shaping acquisition strategies. Some buyers may actively pursue targets with advanced AI capabilities or valuable proprietary datasets, viewing these as catalysts for competitive advantage. Others may look to invest in businesses positioned to benefit from growing demand for computing infrastructure, such as data centres and energy providers. At the same time, acquirers could well become more cautious about legacy businesses that have not embraced technological change, recognising that such targets may struggle to remain competitive.
Although there are numerous examples of AI-related concerns affecting the value of software companies, these dynamics are not confined to the technology sector. Spanning industries from retail to financial services, a target’s approach to AI adoption and its vulnerability to AI-driven disruption may become increasingly relevant to its long-term prospects. For buyers, this means grappling with new forms of uncertainty: How sustainable are the target’s AI-related revenues? How quickly might its technology become antiquated? Where does the real value lie — in the software or in the underlying data?
Due Diligence in the AI Era
Where a target's value depends on AI capabilities, due diligence must go further than traditional processes. Not all AI is created equal: What one seller describes as AI might be a sophisticated machine learning platform, or it might be a collection of relatively simple automation scripts. There is a material difference between the two from a valuation perspective, so buyers need to understand exactly what they are acquiring and whether it can deliver the anticipated benefits. This underscores the importance of due diligence. Key areas for diligence are likely to be how the AI capabilities were developed and the extent to which AI is integrated into the target’s products or systems. As part of that, diligence should also consider the AI’s use and access to trade secrets and confidential information, the nature and type of content used by the AI, and how the AI's outputs are used.
Legal due diligence should also address wider regulatory compliance (including data protection obligations) and intellectual property concerns. We anticipate that AI-related acquisitions will attract increasing regulatory focus, particularly in instances where they could be seen to effect national security, reduce competition or entrench data monopolies.
Bridging Valuation Gaps: Structuring AI Deals
Valuing AI assets is inherently difficult. Performance can fluctuate, technology can become outdated quickly and future revenues may be uncertain. Buyers and sellers often have different views on what an AI business is worth, and bridging that gap often requires more complex deal structuring.
Earnouts and equity rollovers are common methods for addressing this, as deferred consideration can be linked to performance benchmarks. Escrow arrangements may also be used to hold back part of the purchase price pending confirmation that the technology functions as expected.
Addressing AI Use in NDAs and Confidentiality Agreements
AI also raises practical issues at the very start of a transaction, when parties exchange and review proprietary confidential information. Recipients of sensitive data may use AI tools in their daily workflows, and there is a risk that confidential information could be uploaded to those systems. These systems may, in turn, use the data to improve their models, meaning that confidential information could theoretically be incorporated into the system.
Perhaps to address this concern, we are beginning to see AI clauses included in non-disclosure agreements (NDAs). In some cases, existing confidentiality provisions may be sufficient. In others, it may be appropriate to permit conditional AI use, subject to safeguards. Highly sensitive transactions, or transactions involving a target business that is an AI business (or highly dependent on AI for its results and performance), may, however, warrant additional guardrails in NDAs.
Contractual Protections for AI-Driven Value
When AI is central to the deal rationale, buyers will want robust contractual protections relating to ownership, training data, compliance with regulations and accuracy of claims about performance and capabilities.
The overall timing of a transaction, including any period between signing and closing, will also be very important. AI assets can deteriorate quickly — whether through degradation of datasets or changes to underlying technology. Should a significant gap between signing and closing appear likely, buyers should carefully consider the extent to which interim covenants requiring the seller to maintain the business in its current state and preserve core technology may have to be balanced against obligations to ensure the technology continues to evolve in response to market demands.
AI in Deal Execution: Enhancing the Transaction Process
AI is also transforming how deals are executed. Private equity firms, venture capital and other acquirers are increasingly leveraging AI to scale and accelerate their own analytics. Some platforms may gain traction among private equity firms because they can extract information quickly and with the level of detail required by investment teams, enabling faster and more comprehensive analysis of potential targets.
Legal advisers have been using technology to assist with due diligence and document review for many years, but recent advances in generative AI (GenAI) have opened up new possibilities. These tools can help summarise large volumes of documents, identify relevant provisions and even flag risks. That said, AI is not a substitute for legal expertise. Outputs need to be carefully reviewed and validated by experienced practitioners who understand the legal and commercial context and can identify technical errors or informational gaps.
Looking Ahead
We anticipate that AI will continue to redefine the M&A landscape in the years ahead. Businesses that look to better understand these imminent opportunities and risks, and work with advisers who can help them navigate both, will be best placed to succeed.
*Eleanor Bines, a Corporate trainee in our London office, contributed to this advisory.
