AI Due Diligence in Italy: Bringing AI risk and value creation into the investment case

Industry insights

AI Due Diligence in Italy: Bringing AI risk and value creation into the investment case

By Luca Fratesi

28.09.2026 – 6 min read

AI is becoming an increasingly important factor in M&A underwriting. It can affect revenue growth, margins, customer retention, competitive positioning, and the credibility of management's business plan.

Yet AI is still often assessed indirectly in Italian transactions. Commercial Due Diligence covers market dynamics and competition. When done, Technology Due Diligence assesses architecture, scalability, and technical debt. AI implications often sit between the two.

The obvious challenge is that AI DD can easily be perceived as a crystal-ball exercise, given the pace of change in AI and a typical PE holding period of several years. The role of diligence is to bring rigor to that uncertainty.

This year OMMAX has conducted two of the first dedicated AI Due Diligence assessments in large transactions in the Italian market. Combining transaction experience with enterprise architecture, data, AI engineering, and implementation expertise allowed us to translate broad questions about AI into implications for the investment case.

In our experience, a rigorous AI DD needs to cover three areas.

1. Understand AI market opportunities and risks

AI risk is highly sector-specific.

In some industries, the main opportunity comes from automating workflows, improving service delivery, and creating a better customer experience. Education provides a good example, with potential applications across content creation, content production, personalization of the customer journey, and learner support.

In software, AI can have a more fundamental impact on competitive positioning. It can change product propositions, automate existing workflows, enable customer self-build, and shift control of the customer interface.

For software and other technology-enabled businesses, AI can also change where value sits in the value chain. One question we have assessed is who will own the workflow as agentic AI develops: an incumbent may retain a strong position through control of authoritative data, permissions, transactions and workflow-specific context. In other cases, third-party agents could increasingly own the customer interface and capture part of the economics. Understanding this System of Record vs. System of Action dynamic can therefore be critical to assessing long-term defensibility of Software and tech-enabled businesses.

2. Test AI maturity and value creation potential

The key question is how much of the underlying workflow AI can actually perform, from providing suggestions to executing tasks across systems with limited human intervention. Assessing this at the workflow level gives investors a much clearer view of current AI maturity and the remaining automation potential.

The next step is translating maturity into value. For internal use cases, theoretical productivity is only one input. Adoption, implementation effort, and process redesign determine the actual impact. Investors should be conservative when translating hours saved into EBITDA. For customer-facing AI, the focus shifts to adoption, pricing, usage intensity, cost economics, and differentiation.

In one recent AI DD, we assessed current AI implementation, pipeline credibility, and quantified opportunities, combining management sessions with direct input from OMMAX AI engineers to challenge the assumptions behind individual use cases.

3. Assess whether the company can execute

AI roadmaps ultimately depend on the foundations underneath them. This assessment benefits from direct engineering experience. Evaluating whether an AI roadmap can move from pilot to production requires a detailed understanding of architecture, data, models, and orchestration.

AI DD needs to reach the investment case

The output should ultimately inform underwriting.

Depending on the target, AI can affect:

new customer acquisition and churn;
pricing and AI monetization;
productivity and margin improvement;
competitive risk;
required technology investment; and
the value creation plan.

In one recent engagement, we mapped AI risks against individual revenue drivers in the management business plan, assessed the potential impact per driver, and translated this into an AI risk-adjusted ARR and EBITDA scenario.

Five lessons from recent AI DD work

Look at workflows, not the number of AI initiatives: maturity can vary significantly within the same company.
Assess opportunity and risk separately: a credible AI roadmap can coexist with meaningful disruption risk.
AI capability is only one component of defensibility: domain expertise, data, workflow ownership, regulation, switching costs, and GTM can be equally important.
Be conservative when translating productivity into EBITDA: technical automation potential and realized savings are different measures.
Bring the findings into underwriting: where AI affects management assumptions, the impact should flow through to the business case and value creation plan.

Bringing rigor to AI underwriting

AI DD cannot predict exactly how the market will look in five years. It can establish which assumptions matter, test them with evidence, and quantify their potential impact. Doing that well requires a multidisciplinary team combining transaction expertise with enterprise architecture, data and hands-on AI engineering capabilities.

The Italian M&A market is still early in adopting this discipline. Based on two of the first dedicated AI DDs conducted in large Italian transactions, we expect Tech & AI Due Diligence to become an increasingly important part of M&A underwriting as AI becomes more material to investment cases.

OMMAX supports investors and portfolio companies across AI strategy, technology readiness, and AI Due Diligence, helping organizations understand where they stand today and build a credible path to AI-driven value creation. To discuss your AI Due Diligence needs or explore our framework, please get in touch with our team.