Private Equity AI Due Diligence Checklist (2026): 47 Questions for Any Portfolio Company
AI due diligence is no longer optional. Whether you're evaluating an acquisition target or conducting annual portfolio reviews, the ability to assess AI maturity — and translate it into EBITDA opportunity — is now a core PE competency. This checklist covers 47 questions across 6 domains that our operating partners use on every engagement.
Why AI Due Diligence Now?
In 2023, "AI readiness" was a nice-to-have on the due diligence checklist. In 2026, it's a valuation driver. PE firms acquiring a $50M EBITDA business without assessing AI exposure are leaving 3–8% margin on the table — and potentially paying a premium for a company that will face structural labor cost disadvantages within 18 months.
The firms winning on AI don't have bigger technology budgets. They have a structured diagnostic process that identifies the highest-ROI automation and AI opportunities — and integrates them into the 100-day plan.
How to Use This Checklist
This checklist works in two modes:
- 1.Pre-LOI screening: Use domains 1–3 (Data Infrastructure, Current AI Footprint, EBITDA Leverage) in management presentations and data room review. Flags red flags before you're committed.
- 2.Post-close 100-day plan: Use all 6 domains on owned portfolio companies to build the AI roadmap and identify first wins for the investment memo.
Score each item: Green (in place), Yellow (partial/in progress), Red (missing/unknown). Any domain with 3+ reds is a value creation priority.
Domain 1: Data Infrastructure & Quality
AI is only as strong as the data it runs on. These questions expose whether the company has the foundation to support AI deployment — or whether data infrastructure needs to be the first investment.
Is operational data centralized in a warehouse/lake, or siloed across ERP/CRM/point systems?
What percentage of core data is clean, labeled, and accessible to analytics teams?
Does the company have a defined data governance policy with an owner?
Are there active data pipelines refreshing key business metrics (daily or better)?
Has PII/PHI been identified, classified, and access-controlled?
Is the company running cloud infrastructure (AWS/Azure/GCP) or on-prem legacy?
Are there data quality SLAs or known gaps affecting reporting accuracy?
Does IT have a documented data dictionary or catalog?
Domain 2: Current AI & Automation Footprint
Understanding what's already deployed separates genuine AI momentum from vendor slide-deck claims. Most mid-market companies are further along than they think — but not where they claim to be.
What AI or ML tools are currently deployed in production (not POC)?
What workflows are partially or fully automated today?
Is there a dedicated ML/AI engineer or data science team?
What percentage of manual work (by labor cost) has been eliminated through automation in the last 2 years?
Are AI tools bought as SaaS (e.g., Gong, Salesforce Einstein) or built in-house?
Is the company paying for duplicate SaaS tools that overlap with AI capabilities?
Are there known shadow IT AI deployments not sanctioned by IT leadership?
Domain 3: EBITDA Leverage Potential
These questions translate AI readiness directly into EBITDA impact. Each 'no' is a specific value creation lever for the ownership period.
What are the top 3 labor-intensive processes consuming >5 FTEs each?
Where does the company experience the highest error rates (billing, fulfillment, QC)?
What is the customer churn rate, and is churn prediction in place?
Does the company have an active demand forecasting capability?
Are there manual reporting/finance processes consuming >20 hours/month?
What is the average cost per customer acquired, and is propensity modeling in use?
Has the company benchmarked AI adoption against sector peers?
What is the current SG&A as a % of revenue vs. industry benchmark?
Domain 4: Technology Stack & Integration Risk
Legacy tech stacks aren't a dealbreaker, but they dictate the speed and cost of AI deployment. Identify integration risk before the 100-day plan commits to timelines.
What is the core ERP/CRM system, and how old is the implementation?
Are there open APIs/webhooks on core systems that support AI integration?
What is the average age of on-prem infrastructure that would require modernization?
Has a technology debt assessment been performed in the last 18 months?
Are there contractual restrictions (vendor lock-in) that limit AI tool adoption?
What is the IT budget as % of revenue vs. industry peers?
Is there a documented integration architecture or ESB (enterprise service bus) in place?
Domain 5: Talent & Organizational Readiness
Technology without organizational readiness delivers theoretical EBITDA, not real EBITDA. Talent and culture are the constraining variable in most AI transformations.
Does the management team have a defined AI strategy or roadmap?
Is there a CTO, VP of Engineering, or Chief Digital Officer in the org chart?
What percentage of employees have completed any AI literacy training?
Has the company run any AI pilot programs in the last 12 months?
Are there incentive structures aligned to digital transformation outcomes?
Does the company have relationships with AI vendors or consultants?
What is current employee resistance level to automation-driven change?
Domain 6: AI Governance & Risk
AI governance is now table stakes for regulated industries — and increasingly for any institutional investor. These risks compound with scale.
Is there a formal AI/ML usage policy or acceptable use framework?
Have model risks (bias, hallucination, drift) been assessed for any deployed AI?
Are AI model outputs auditable and explainable to regulators/auditors?
Is there a data breach response plan that covers AI system compromises?
Does the company operate in a regulated industry (HIPAA, SOX, GDPR) with AI implications?
Have third-party AI vendors been vetted for SOC 2 or equivalent compliance?
Is there a process for monitoring AI model performance post-deployment?
Are AI-generated outputs (e.g., customer communications, financial projections) reviewed before use?
Interpreting Your Scores
Company has strong foundations. Focus due diligence on identifying the highest-ROI deployment opportunities within existing infrastructure.
Material gaps exist but the core business is sound. AI roadmap should start with data infrastructure and 2–3 high-ROI automation wins.
AI creates risk, not uplift, in near term. Price in 18–24 months of remediation. Consider whether deal thesis depends on AI EBITDA.
Automatic Red Flags (Regardless of Total Score)
Any of these answers should trigger deeper diligence or deal structure adjustment:
No clean, accessible data for the company's primary revenue driver
AI tools in production with no monitoring or performance review process
Regulated industry (healthcare, financial services) with no AI governance policy
Management team that cannot articulate the company's AI roadmap or competitive position
Active litigation or regulatory inquiry related to data, privacy, or automated decision-making
Vendor lock-in on legacy systems with no modernization path budgeted
Get a Structured AI Assessment in 10 Days
Our operating team runs this full 47-question assessment, cross-references benchmarks, and delivers a board-ready report with quantified EBITDA opportunities. Turnaround: 10 business days.