The message from corporate counsel and private practice leaders is clear: AI is already reshaping due diligence in M&A, including speeding reviews, standardizing outputs, shifting budgets. But, as Sam Dorman (Datasite) recently discussed with Alex Haskell (Salesforce Ventures) and Ella Sharpley (Travers Smith), the firms winning with AI aren’t just using tools; they’re systematizing how they use them, aligning expectations to use cases, and keeping a human layer on judgment and risk.

Delivering real gains and why adoption differs

AI is already reshaping core stages of the M&A diligence workflow – not by replacing legal judgment but by accelerating the data-heavy tasks that traditionally slow teams down. The biggest benefits today come from three clear use cases, each offering measurable efficiency gain, while adoption patterns differ between in‑house teams and private practice based on accuracy expectations, volume, and workflow design.

The areas where AI is delivering the most impact:

1. Rapid analysis of public data before diligence even opens

Early-stage buy‑side teams are using chat‑style AI to synthesize publicly available information, such as Companies House filings, financial accounts, or website content, into structured, decision‑ready insights. A common example is generating a clean, accurate cap table (including share classes, voting rights, and economic rights) from messy filings and articles. What once took hours now takes minutes, with lawyers applying a quick sense check.

2. Playbooks for consistent, high‑volume document review

Firms are building standardized prompts (‘playbooks’) for repetitive asset classes: leases, commercial contracts, IP registers, and more. AI extracts the key commercial and legal terms, producing a consistent rider for the diligence report. Lawyers then verify accuracy and materiality, but the manual review load drops dramatically.

3. Full‑stack first drafts for minority or high‑volume investments

High-frequency investors are already uploading entire data rooms to leading LLMs and instructing them to produce diligence reports in their own preferred format, guided by prior reports. With a 95-98% accuracy threshold accepted in these contexts, teams are meaningfully reducing external spend and redirecting that work internally through AI‑assisted processes.

Why adoption looks different in-house versus private practice:

• In‑house teams operate at higher volume and often have more flexible accuracy thresholds. making ‘95% correct’ first drafts both acceptable and valuable. AI enhances throughput and frees teams to focus on strategic issues rather than extraction.
• Law firms, by contrast, work to a 100% accuracy expectation on high-stakes transactions. That creates a hybrid model: AI accelerates the review and extraction, while human lawyers validate the output and provide the ‘so what’ analysis that informs negotiation, pricing, and risk allocation.

The challenges and how M&A teams are solving them

As AI moves deeper into the M&A workflow, the biggest hurdles aren’t technical; they’re operational. Teams are learning that effective adoption depends on clarity, consistency, and trust. The firms making the fastest progress are the ones building repeatable processes around AI rather than treating it as an ad‑hoc tool.

Core challenges and how to overcome them:

1. Get specific with prompts

“Give me a diligence report” isn’t enough. Strong outputs come from:

• Clear, patterned prompts grounded in prior reports you actually rate
• Lists of the top issues you always evaluate
• Role-based framing (e.g., “Act as buy-side counsel focusing on X/Y/Z risks”)

2. Prioritize use cases that are worth systematizing

Playbooks take real time to build, so focus on high-volume, repeatable work where standardization compounds value. Typical candidates include leases, customer/vendor contracts, cap tables, IP schedules, HR policies, and pensions indicators.

3. Institutionalize learning across the team

Regular show-and-tell sessions (“Who used AI this week? What worked?”) prevent siloed experimentation and accelerate collective progress. Junior lawyers often spot the most automate-able tasks, so make space for them to lead.

4. Build trust through evidence

Side-by-side comparisons show AI is often as accurate as humans (and sometimes more so) for extraction and summarization. Keep a human layer for materiality, nuance, and negotiation positioning, as this is where judgment truly matters.

5. Navigate shifting budgets, fees, and value expectations

Fee pressure on commoditized diligence tasks is already increasing, especially as in-house teams face mandates to “use the AI we’re paying for.” External counsel still win when they deliver:

• Verified accuracy that avoids false comfort
• Judgment on what matters and why
• Deal strategy: turning findings into SPA (share purchase agreement) positions, issues lists, and integration actions

In short, AI may compress the cost of extraction, but it doesn’t diminish the value of interpretation. The firms that balance efficiency with expert judgment will retain their edge.

What’s next and how M&A teams can prepare

Over the next 12-24 months, AI is expected to move from ‘pilot use cases’ to deeply embedded, end‑to‑end workflows across the M&A lifecycle. The biggest changes will come from AI shifting where diligence happens, how first drafts are produced, and which parts of the process remain human‑led. To stay ahead of the curve, deal teams should pair an understanding of what’s coming with some practical steps they can take now to futureproof their processes.

Key developments to expect and how to prepare:

1. In‑platform diligence

AI agents will sit directly inside the VDR, eliminating download-upload steps and enabling secure, role-aware analysis with full audit trails.

2. Auto‑drafting beyond summaries

Expect increasingly reliable first drafts of heads of terms, SPA issues lists, and clause-level redlines tailored to your playbook positions.

3. Bifurcation of work

High-stakes ‘elite’ M&A will stay human-led with AI support, while everything else becomes more commoditized, faster, and more AI-driven.

4. Codify your playbooks

Prioritize your top three document categories (e.g., leases, key customer contracts, employment/benefits) for standardized prompts and outputs.

5. Create a model pack

Build a library of your best prior reports so AI tools consistently mirror your firm’s preferred structure, tone, and risk priorities.

6. Set acceptance thresholds

Align accuracy expectations with deal type, for example, 95-98% for minority growth deals versus full verification for critical buy‑side transactions.

7. Assign ownership and track improvements

Nominate owners for prompt libraries and weekly AI learnings; measure time saved and accuracy rates to compound improvements.

8. Keep the ‘so what’ human

AI can extract, summarize, and structure, but translating findings into positions, price adjustments, or protections remains a human value-add.

AI won’t replace expert deal judgment, but it will redefine who wins on speed, consistency, and cost. The firms that industrialize their approach now will set a new standard for diligence quality and velocity.

Insight. Innovation. Impact.

Take the next step. Schedule a demo and experience the tools reshaping the dealmaking landscape.