Artificial intelligence has moved quickly from a technology talking point to something that now sits at the heart of merger and acquisition transactions. Across all sectors and deal sizes, lawyers are seeing AI influence target identification, diligence, value assessment by the potential buyer, and risk allocation.
At the same time, consideration of artificial intelligence is no longer limited to technology-sector deals or “AI companies” in the narrow sense. Instead, it is increasingly part of the standard analysis on any prospective deal that involves a company’s use of AI.
Curtis Cusinato, a partner in the Toronto office of Bennett Jones LLP specializing in corporate and securities law with a focus on cross-border mergers, states, “AI is everywhere in M&A now.” Its influence stretches across the full lifecycle of a transaction, from origination through to post-closing integration.
Curtis Cusinato
For Cusinato, one of the most visible changes he sees is the rapid and widespread adoption of AI in deal execution. AI tools are now routinely embedded in sourcing platforms and data rooms, but their real impact is most obvious in diligence. “Data sites powered by AI can streamline diligence,” he explains, allowing legal teams to process and analyze vast volumes of documentation that would previously have required significant manual review.
That shift is not just about efficiency. It is also changing how legal work is allocated. AI now supports “diligence questions, research, and document drafting,” Cusinato notes, which has reshaped the role of junior lawyers. Instead of spending most of their time on first-pass document review, he says, they are increasingly validating outputs, testing assumptions, and focusing on higher-value risk analysis.
One of the most tangible examples, he notes, is in contract review. AI tools can now scan thousands of agreements and benchmark provisions in purchase-and-sale agreements against large datasets of precedent transactions. Clauses can now be quickly assessed as “seller friendly,” “buyer friendly,” or “market standard” – something that previously required significant manual benchmarking.
Cusinato also highlights how AI is changing the way lawyers think about using prior transactions to frame the deals they are working on. Increasingly, he says, AI is used to review “publicly available documentation” and summarize how counterparties have behaved in past transactions. In some cases, it can even analyze a lawyer’s prior deal history, identifying patterns in negotiated positions and effectively mapping a data-driven negotiation profile.
Despite these advances, Cusinato emphasizes that AI has not replaced lawyers' core function in M&A. If anything, it has sharpened it.
As for determining the value that can be attributed to AI in companies that are adopting it, Cusinato says the central focus for lawyers remains on ownership rights and how “pristine” the data is – especially in cases “where datasets, training inputs, and licensing rights are involved.”
That focus extends to familiar areas of potential risks: confidentiality, privacy, intellectual property infringement, and professional responsibility. Cusinato points to “confidentiality restrictions, professional ethical restrictions, privacy concerns, and infringement of intellectual property” as recurring issues that become more acute when AI systems are embedded in a business.
Another key concern is reliability. Lawyers are increasingly required to assess “the quality of outputs” generated by AI systems, particularly with well-documented risks such as hallucinations. That has made validation a critical part of modern diligence, especially where AI-generated insights influence commercial decisions.
As a result, third-party experts are now a more common feature of AI-related transactions. Curtis notes that buyers frequently engage specialists to stress-test systems and verify whether claimed capabilities are genuine or simply “glorified SaaS software” marketed as AI innovation. This external validation layer has become a standard feature of sophisticated deal processes.
Konata Lake, a partner at Torys LLP who focuses on M&A, venture capital, and private equity investments, approaches AI in transactions from both valuation and structural perspectives. He too emphasizes that AI is now relevant across virtually all sectors, not just technology. “At its core, an M&A transaction is a buyer deciding that a target company has assets, rights, products, or services they want to acquire,” Lake explains. “The question now is how AI affects the value of those assets and future opportunities.”
Konata Lake
A central issue is data. Buyers are increasingly focused on whether companies are using AI to unlock value from existing datasets or generate new revenue streams. That raises foundational legal questions about consent, ownership, and usage rights.
“We need to understand where the data came from, whether the company has the legal rights to use it, and whether the proposed uses align with what users originally consented to,” Lake says.
Privacy is especially sensitive in regulated industries such as healthcare, where data use is heavily scrutinized and regulatory expectations continue to evolve. “One of the challenges with AI is that the regulatory landscape is evolving quickly,” he notes. “You have to consider not only the rules that exist today, but also what future regulations may look like.” These factors can have a significant impact on how the target company uses AI and how it is valued by the potential buyer.
AI is also reshaping how buyers think about “disruption risk.” Rapid advances in large language models may put traditional software businesses under unexpected competitive pressure. As Lake puts it, “A buyer now has to ask whether the company they’re acquiring could become obsolete if a major AI platform releases a competing feature.”
Lake also acknowledges that law firms themselves are increasingly using AI tools for research, precedent analysis, drafting, and contract review, but he is clear that these remain assistive technologies. “It’s still a tool that assists lawyers rather than one that replaces legal judgment,” he says.
M&A teams are also becoming more multidisciplinary, he points out, with privacy, IP, employment, tax, and regulatory specialists playing a more integrated role in AI-related deals.
For Lake, the broader challenge is timing. “Businesses need to modernize and invest in technology,” he says, “but they’re doing so in an environment where what’s cutting-edge today may be outdated very quickly.”
Andrea Johnson, who leads the national corporate group at Dentons Canada LLP, sees many of the same shifts in how AI is considered, particularly in how buyers evaluate targets. AI, she notes, is no longer confined to technology companies. “Even businesses in traditional industries are starting to experiment with infusing AI through their operations,” she says. Companies are using it to drive efficiency, streamline workflows, and, in some cases, create entirely new revenue streams.
Andrea Johnson
That change is reshaping deal strategy. Increasingly, transactions involve a “buy-versus-build” analysis. As Johnson puts it, “Can we develop these capabilities internally, or should we acquire a company that’s already further along in its AI journey?” That question is now part of mainstream M&A decision-making.
Much of the value, Johnson adds, is not in model development itself but in the application layer – “companies that use AI to unlock operational efficiencies or create new products and services.” As a result, diligence has evolved accordingly.
What Johnson describes as a “dashboard approach” to valuation is becoming common, where acquirers assess AI maturity on a spectrum: “red, yellow, or green in terms of AI adoption – whether they’re not using it, experimenting with it, or actively generating measurable business benefits.”
Importantly, AI adoption alone is not enough to justify valuation premiums. “Buyers want to see whether AI adoption is actually increasing revenue, reducing costs, or improving profitability,” she says. In other words, impact – not experimentation – drives value.
At the same time, expectations are still catching up to reality. Johnson describes the market as emerging from a “trough of disappointment,” where many organizations invested heavily in AI tools without seeing immediate returns. More recently, however, early adopters have begun to demonstrate meaningful operational gains through automation, AI agents, and redesigned workflows.
That progress has brought new legal considerations into diligence. AI-related transactions now routinely involve assessments of data governance, IP ownership, privacy compliance, and regulatory exposure. Increasingly, Johnson says, companies are implementing formal governance frameworks, including the National Institute of Standards and Technology (NIST) Risk Management Framework. It’s a voluntary guidance document designed to help organizations design, develop, deploy, and evaluate trustworthy AI systems while mitigating negative impacts. Using ISO-aligned standards can also ensure that AI is deployed transparently and ethically, mitigating risks such as algorithmic bias, data misuse, and a lack of accountability.
“Buyers want to see that companies are not just using AI, but using it safely and thoughtfully,” Johnson notes.
Law firms are adapting in parallel. AI tools now allow lawyers to analyze entire populations of contracts rather than relying on sampling. “In the past, lawyers might review only the most important contracts and a sample of the rest,” Johnson says. “Now… AI allows us to analyze the entire population quickly and identify trends, risks, and inconsistencies.”
For Johnson, the result is a fundamental shift in client expectations: faster analysis, broader coverage, and more data-driven insight at every stage of a transaction.
Garrett Hamel, an Ottawa-based partner with Gowling WLG, says AI has become a powerful tool for lawyers in deal diligence. “AI has become a practical enabler of faster and more comprehensive review processes,” he says, particularly in data-heavy transactions involving thousands of contracts. Where sampling once defined the limits of review, AI now enables full-population analysis, surfacing issues that might have gone undetected previously.
Garrett Hamel
However, he adds, that capability has shifted client expectations. Buyers now expect not just faster diligence, but deeper insight. Hamel describes this as a shift toward a more value-driven model, where lawyers are expected to identify patterns that affect pricing, structure, and risk allocation – not just flag issues in isolation.
He also emphasizes how AI is reshaping benchmarking. AI tools now allow lawyers to compare contractual provisions against large datasets of prior deals in real time, improving consistency and bringing a broader market perspective into negotiations.
Duncan Snyder, Hamel’s colleague at Gowling WLG, who is also a partner in the firm’s corporate and private M&A group and based in the firm’s Waterloo Region office, notes how AI is influencing valuation discipline and behavioural risk in transactions. One of the key challenges, he says, is distinguishing real capability from “AI branding” that does not translate into operational value.
Duncan Snyder
He says buyers are increasingly asking whether a target is truly defensible in an “AI-enabled market,” rather than simply part of the AI narrative on how disruptive the technology is.
That concern has led to more detailed AI-specific representations and warranties, particularly around data use, model integrity, and system governance. Snyder sees this as an area of M&A transactions that, like the impact of earlier waves of technology, will continue to evolve as AI adoption and integration across industries advance.
Internal use of AI is another growing risk area. Snyder highlights concerns about employees “inputting sensitive or confidential information into external AI tools,” which can create privacy, IP, and regulatory exposure depending on the sector.
He also underscores the importance of human oversight. In one instance, he notes that AI-assisted tools misinterpreted source documents – for example, by incorrectly reading a handwritten date – leading to downstream analytical errors. While caught by lawyers, the incident illustrates the ongoing need for professional review.
Hamel adds that foundational diligence around data security and ownership is becoming central. If a company cannot clearly establish its rights to the data feeding its models and systems, it can materially impact valuation or even derail a transaction.
Both also flag the risk of overpayment. AI enthusiasm can accelerate deal timelines, but also inflate expectations. Buyers must therefore ground valuations in demonstrable use cases and measurable outcomes rather than projected capabilities.
Even so, AI is clearly accelerating transactional workflows. Hamel points to real-time benchmarking tools that significantly increase efficiency and consistency in drafting and negotiation. The trade-off, he notes, is that faster execution can compress the informal relationship-building that often supports successful dealmaking.
Still, both he and Snyder return to a consistent theme expressed by lawyers who work on deals involving companies that are using AI: As Snyder puts it, legal work in this space remains anchored in “risk identification, mitigation, and judgment,” even as the tools continue to evolve.