AI Due Diligence: What Every Investor Should Ask
August 31, 2026
Author - NSB&Co
An impressive AI demonstration does not automatically translate into a valuable or investable business.
A live demo may show technical potential, but it does not by itself establish that the business has the rights, controls, economics, and operational resilience required for commercial deployment. Weak governance, unclear data rights, hidden infrastructure costs, and dependence on external providers can erode value as quickly as AI may create it.
The central diligence question is not simply whether a company uses AI. It is whether the company has durable, documented, and commercially usable rights to develop, operate, scale, and, where relevant, transfer that AI-enabled capability.
Before investing in or acquiring an AI-enabled business, due diligence should examine:
- AI assets and architecture: What proprietary, open-source/open-weight, and third-party models, code, datasets, cloud services, and APIs support the product? What licence terms, usage restrictions, attribution obligations, and provider dependencies apply?
- Data and intellectual-property rights: Does the business have documented rights and permissions to collect, access, process, retain, train on, and commercialise the data, models, and outputs involved in its intended use case?
- Transferability: Are critical licences, data rights, customer agreements, and vendor contracts transferable in the proposed transaction, or subject to assignment restrictions, consent requirements, or change-of-control provisions?
- Commercial viability: Does the business model remain viable after accounting for compute, inference, API, cloud, security, monitoring, maintenance, and human-oversight costs? A model that works economically at pilot scale may not do so at production scale.
- Performance and scalability: Has the system been tested in production or production-like conditions at expected volumes, latency, reliability, security, and cost thresholds—not solely in a controlled demonstration?
- Governance and accountability: Are documented oversight, monitoring, testing, escalation, accountability, and incident-response mechanisms in place?
- Third-party concentration: Could a critical vendor's price changes, service interruption, model retirement, usage limits, policy changes, or contractual restrictions materially affect the business?
- Deal implications: Are identified risks sufficiently material to justify a valuation adjustment, targeted indemnities, escrow, pre-closing conditions, transition arrangements, or a defined remediation plan?
The gap between demonstration and deployment is often where material transaction risks emerge. A demo will not necessarily reveal limitations around data provenance, IP and contractual rights, provider concentration, compliance controls, unit economics, or the cost of operating the system reliably at scale.
How N.S. Bhargava & Co. Can Help
At N.S. Bhargava & Co., financial, operational, compliance, governance, and risk considerations can be incorporated into the broader due diligence process for AI-enabled businesses.
Where specialised technical, cybersecurity, data-protection, or legal analysis is required, these matters can be evaluated in coordination with appropriately qualified experts. This gives investors a more connected view of transaction risk, rather than treating AI as an isolated technology issue.
Explore our capabilities:
Due Diligence & Certification | Audit & Assurance | Risk Advisory