Summary
Summary
- AI-washing means presenting ordinary software as AI-driven to win a higher valuation. In 2026, buyers test for it as a standard part of due diligence.
- The core question is simple: if the underlying model were swapped for a competitor’s tomorrow, would the product’s value survive? If yes, the AI is a feature and the company is priced as core SaaS.
- Defensible AI value comes from proprietary data, workflow depth, measurable outcomes, unit economics that hold up after inference costs, and customer pull that predates the AI launch.
- Software companies with strong fundamentals lose nothing from this audit. Clean retention, margins and evidence of AI-driven outcomes are what earn a premium now.
01 · Research
From AI label to AI audit
The market stopped paying for the word.
Through 2024 and much of 2025, adding a generative AI feature to a software product was often enough to change the conversation with buyers and investors. Model APIs made it quick to ship an assistant, a summariser or a copilot, and the label carried weight in its own right.
By 2026 buyers have seen hundreds of those launches. The same API is available to every competitor, so a feature that anyone can build in a sprint does not justify a structural premium. Disclosed AI deal value in Q3 2026 moved decisively towards infrastructure, as our Death of the Tech Premium analysis shows.
02 · Research
The five tests buyers apply
A practical AI-washing audit.
- 01The model-swap testCould a competitor reproduce the AI capability by calling the same third-party model with a reasonable prompt? If so, buyers treat it as a feature and price the company as core SaaS.
- 02The proprietary data testDoes the product learn from data the company has the rights to and competitors cannot get: labelled workflows, outcomes, transaction histories? Data rights and provenance are checked line by line.
- 03The workflow-depth testIs the AI embedded where work actually happens, with integrations, permissions and audit trails, or bolted on as a chat box? Deep workflow embedding drives retention that survives model changes.
- 04The outcome testCan customers show measurable results such as hours saved, errors reduced or revenue added? Buyers increasingly ask for customer-level evidence, not marketing claims.
- 05The margin testWhat is gross margin after model and inference costs? AI features that consume a large share of revenue in API fees are valued on the true margin, not the headline one.
These sit alongside the standard diligence work in our AI due diligence guide and quality of revenue framework.
03 · Research
How the audit changes the multiple
Two pricing paths.
| Finding | How buyers price it | What moves the number |
|---|---|---|
| AI is a feature on a strong SaaS core | Core SaaS multiple, based on growth, retention and margin | Retention and cash conversion |
| AI drives outcomes on proprietary data | Premium to core SaaS, justified by defensibility | Data rights, outcome evidence, workflow depth |
| AI is a thin wrapper with high API costs | Core SaaS multiple applied to post-inference margin | Reducing inference cost, adding proprietary signal |
The audit also feeds directly into EBITDA normalisation. Buyers adjust for capitalised AI development, one-off model costs and promotional AI pricing that will not last. Our EBITDA add-back standard sets out which adjustments survive diligence. For current benchmark ranges see SaaS valuation multiples 2026.
04 · Research
How to come out on the right side
Clarity earns the premium.
- Split the revenueShow AI-attributable revenue and usage separately from core revenue, with the method stated.
- Report post-inference marginsPresent gross margin after model costs, plus your plan to reduce them through caching, smaller models or routing.
- Document the dataList proprietary datasets, how they were collected, and the contractual rights to train on them.
- Bring outcome evidenceTwo or three customer case studies with hard numbers outweigh any amount of positioning.
A strong software business loses nothing in this audit. Good retention, healthy margins and honest AI metrics are exactly what serious buyers are looking for, and they reward them.
Reference
Frequently asked questions
What is AI-washing in M&A?
Presenting a software product as AI-driven, or overstating how much of its value comes from AI, to support a higher valuation. Common examples are thin wrappers around a third-party model API with no proprietary data, evaluation or workflow behind them.
How do buyers value AI software companies in 2026?
On the same fundamentals as other software (growth, retention, margins and cash conversion), with a premium only where the AI creates defensible value: proprietary data, deep workflow integration and measurable customer outcomes that a competitor could not replicate by calling the same model.
Do inference costs affect valuation?
Yes. Buyers recalculate gross margin after model and inference costs. A product whose AI features consume a large share of revenue in API fees is valued on its true margin, not its pre-AI margin.
How can a founder avoid being discounted as AI-washing?
Document what is proprietary, show outcome data from customers, report margins after inference costs and separate AI-attributable revenue from core revenue. Clarity removes the discount buyers apply to uncertainty.
Keep reading
Related analysis
Research · Research
Quality of Revenue Framework
An institutional framework for scoring revenue quality in digital asset transactions.
Research · Cross-sector
The 2026 EBITDA Add-Back Standard: 20 adjustments and how buyers treat them
Every seller adjusts EBITDA. Every buyer's quality of earnings team unpicks it. This is the reference table we use on live mandates: 20 common add-backs, the evidence that gets each one accepted, and where the line sits.
Research · AI
Inside the Agentic Stack: Why Supabase and Salesforce Are Buying AI Infrastructure, Not AI Features
AI agents create databases, call models and talk to customers at machine scale. The acquirers winning the infrastructure race are buying the layers underneath those agents, not another chatbot on top.
Explore the sector:SaaS acquisitions
About the analyst
Joash Boyton
Founder and Managing Director, Acquiry · Melbourne, Australia · Global coverage
Joash Boyton is the Founder and Managing Director of Acquiry, a specialist M&A advisory firm focused on the acquisition and sale of businesses. He executes buy-side and sell-side mandates from USD $1M to $500M across technology, SaaS, fintech, payments, gaming, blockchain and emerging verticals, and is not limited to them. Any sector, any market.