Contents
A clear signal of the global trend toward using artificial intelligence (AI) in the conduct of mergers and acquisitions (M&As) is the release by Anthropic in late February of plugins that support core deal tasks such as sourcing and screening potential M&A prospects and running due diligence checklists.
Tools like this illustrate the direction of travel: AI is moving beyond generic productivity gains into purpose-built functionality that can reshape how quickly and how widely acquirers can operate.
In the New Zealand M&A market, we expect that, as AI becomes standard practice, as well as lifting the volume of transactions it will also give rise to new legal and governance considerations for boards, buyers and sellers.
Changes to M&A workflows
Faster, broader target assessment
AI tools can analyse rapidly the target's financials, customer and contract information, and market/benchmark information to generate an initial view of how viable a target is. The boost is significant in both speed and scope, and the time saving is likely to mean that:
- buyers can assess more opportunities (including smaller, less acquisition-ready businesses); and
- new potential acquirers may enter the market because an AI-assisted process reduces the amount of in-house time and effort required.
AI-supported due diligence (often deeper, not just faster)
The depth of due diligence is also likely to increase, as AI is increasingly used to sort and analyse document-heavy diligence, allowing for broader coverage, at a faster pace and producing helpful outputs (e.g. key clause extraction, issue lists, deviation reports, risk scoring).
As AI assists due diligence to become broader and faster, flow-on effects emerge. Deeper diligence reduces unknowns, which could result in buyers identifying more risks and pushing harder on indemnities, conditions precedent, and price adjustments in response.
Likely impacts on the NZ M&A landscape
Shorter timelines - especially early-stage
AI may have the greatest impact in the initial assessment and first-pass diligence phases, where deals often slow down or stall due to limited resources and a flood of information. The ability to complete this early-stage work more quickly could also reduce timelines from teaser, to indications of interest, to exclusivity, which would consequently reduce the time available for sellers to "clean up" issues once a process begins. It may also grow the pool of potential targets, as time and cost savings allow for assessment of opportunities that might previously have been ruled out.
Increased competition (and potentially more cross-border interest)
Cheaper, faster assessment helps established acquirers run more processes concurrently, while also lowering the bar for resource-constrained and newer acquirers to explore M&A. This could lead to more buyers competing for deals. It may also make NZ targets easier for offshore buyers to identify and prioritise particularly in sectors where NZ is well-regarded, such as agribusiness and SaaS.
More complex acquisition strategies
AI may accelerate acquisition strategies by making bolt-ons easier to identify and quicker to assess. That could also make larger “mostly attractive” acquisitions less necessary or attractive, if buyers can build scale through a series of smaller, more targeted deals over time or even contemporaneously. However, such strategies will still have increased complexity around sequencing, conditionality (including regulatory approvals) and integration.
Legal and governance: what changes when AI is in the loop?
A baseline issue for everyone: accuracy, explainability and human oversight
AI can accelerate processes, but outputs must be checked and tested. The risks are well-known and include hallucinations, overlooked nuance and model drift. The key question is what level of human review is reasonable for the deal - and this will inevitably be assessed with hindsight if a transaction goes wrong.
For buyers: diligence scope, materiality thresholds, and “data hygiene”
If AI allows greater volumes of material to be reviewed, buyers are likely to:
- push for lower materiality thresholds
- seek greater access to operational datasets, and
- expect better “data hygiene” from targets.
This can also cut the other way: buyers may rely on AI summaries without adequate and appropriate checking and source familiarity. That could create risk if the buyer later alleges a breach of warranty where the underlying circumstances leading to the warranty claim were disclosed in the due diligence material provided.
Competition law effects
AI may also help flag competition issues by identifying market relationships and overlaps early. Early awareness could allow buyers to structure deals to manage those risks (for example, through carve-outs or undertakings) or to step away earlier from problematic targets.
Directors’ duties and governance: faster and more frequent deals
AI may increase the volume and pace of acquisition opportunities, which in turn is likely to put more pressure on board processes and oversight, raising governance questions.
- What is a “reasonable” decision-making process in a shortened timeline?
- How should boards challenge and validate management’s AI-supported recommendations?
- What rules, controls and accountability are needed when AI tools contribute to strategic decisions?
Boards will still apply their usual oversight in considering key risks and applying independent judgement when assessing management’s recommendations in respect of a transaction. However, we expect there will be an increased focus on understanding what the recommendation was based on, the nature of the due diligence undertaken to support it and how AI was used in that process.
For sellers/targets: preparedness and record creation
As buyers ask for more data, sellers will need to lift their baseline governance earlier, including around contracts, compliance registers, privacy/cyber incident response, and HR documentation.
A specific emerging issue is the growth of AI-enabled meeting transcription and note-taking tools. Even if the transcripts are not “formal records”, they can contain candid statements, and references to issues that were ultimately never escalated or pursued.
While due diligence is not litigation discovery, sellers should assume that anything they hold and can retrieve may be requested, and that refusal may itself raise questions.
AI as a diligence topic
Increasingly, buyers will diligence not only the target’s financial and legal position, but also its use of AI. This might include:
IP and ownership:
Who owns data, prompts and outputs, and what do vendor terms allow?
Confidentiality:
Has confidential information been uploaded into third-party tools?
Privacy:
Does use of personal data in AI systems comply with the Privacy Act and stated purposes?
Cybersecurity:
Are AI tools creating new attack or data leakage risks?
Overseas regulation:
Do overseas regimes effectively set the standard for customers and counterparties?
Target use of AI should be considered in any deal protections: specific warranties on AI use, data handling, IP, and compliance; targeted indemnities; and conditions precedent.
Impacts on deal mechanics
Warranties and disclosure
Where AI enables faster, more data-heavy due diligence processes, buyers may push for broader warranties. This could also increase friction over the concept of “fair disclosure” where disclosure is extensive and the buyer uses AI to summarise it.
Conditions and sequencing
Where AI enables more complex strategies involving multiple acquisitions, deal documents will need to manage sequencing risk, typically through the use of clear conditionality (including linked completions), robust long-stop and termination rights, and agreed fee/cost allocation if one acquisition does not proceed. It also increases the importance of early planning for any regulatory engagement or approvals (including with the Commerce Commission and Overseas Investment Office.
More information, more speed, more room for misunderstanding
More information and less time may increase the risk of misunderstandings and overreliance on AI summaries. The best mitigant is discipline, and ensuring that AI outputs are treated as working tools rather than final advice.
AI readiness
We expect AI to make M&A faster and more competitive, particularly in the early deal stages. It is likely to increase the number of potential targets that are identified and screened, and could also bring more would-be acquirers into the market by reducing the time and internal resource needed to run a process. That is likely to make for more active markets, but it also raises the bar for governance, document-hygiene, and legal risk management.
For buyers, the opportunity is to move earlier and with greater confidence; the challenge is to ensure judgment and accountability are retained. For sellers, the opportunity is the potential for more, and earlier, acquirer interest; the challenge is being diligence-ready.
The firms and boards that adapt their processes now, including policies, playbooks, and deal documentation, will be better placed as AI becomes a standard part of M&A execution.
