AI and the Recomposition of Tech Value
72% of SaaS M&A targets now reference AI. Median EV/Revenue for AI-native SaaS sits at 12.5x. But the premium is fragile — and the commoditisation trap is real.
72% of SaaS M&A targets now reference AI. Median EV/Revenue for AI-native SaaS sits at 12.5x. But the premium is fragile — and the commoditisation trap is real.
The AI valuation premium is being driven by a genuine recomposition of tech value. Buyers are no longer pricing software on revenue multiples alone. They are pricing proprietary data, contractual moats, regulatory compliance, and net revenue retention. These are the core AI attributes that the CIFSO protocol has formalised.
Proprietary data (I-16 in the CIFSO framework) is the foundation. Datasets that cannot be replicated are the primary source of durable AI value. If a competitor can rebuild your training data from scratch in six months, the asset has no certifiable moat in the AI dimension.
The commoditisation trap is the inverse. 40% of submitted assets present AI features built on thin wrappers around public LLMs — Claude, GPT, Gemini — without proprietary data, contractual depth, or technical differentiation. The valuation premium these assets claim is not supported by the CIFSO assessment.
EU AI Act compliance (S-42) is emerging as a binary filter. Assets that cannot demonstrate compliance with Articles 9–15 face a structural devaluation. Assets that can demonstrate compliance — and have documented it at submission — command a verifiable premium. This is the first regulatory certification arbitrage in European tech M&A history.