The clause that says more than the price

Pascal Daloz, who runs Dassault Systemes, agreed on July 23 to hand over 1.8 billion dollars in cash for ArisGlobal, a vendor most owners outside pharma have never heard of. The number that matters, though, is the one he refused to pay up front. Up to 200 million dollars of the deal, roughly 165 million euros, is held back and released only if ArisGlobal hits multi-year AI-revenue milestones.

That is not how these announcements usually read. The press release leads with a unified AI platform for the life sciences, the language of ambition. The contract underneath it is the language of caution: a large, cash-rich acquirer bolting an earnout onto a deal specifically so the seller has to prove the AI money is real before the last cheque clears.

What Dassault is actually buying

ArisGlobal sells the software that keeps drugmakers on the right side of regulators. Its platform handles pharmacovigilance and safety reporting, and it processes more than 12 million patient safety reports a year for over 200 customers, among them half of the 50 largest biopharma companies in the world. It employs more than 1,300 people and is expected to bring in about 175 million dollars, near 160 million euros, of revenue in 2026.

Dassault already sells modelling and simulation tools used across drug discovery and manufacturing, so the fit is real rather than financial engineering. The plan is to stitch ArisGlobal's compliance and real-world evidence data into that stack and sell one intelligence platform across the whole therapy lifecycle. The base price of 1.8 billion dollars against 175 million in sales works out near ten times revenue, a full price for a business growing inside a regulated niche.

Why the earnout is the signal for your own AI bets

The lesson here is not about pharma, it is about how a serious buyer now prices the word AI. Dassault could have paid the whole sum in cash, and with its balance sheet it barely notices the outlay. It chose instead to separate the part of the value it can see today from the part that depends on AI revenue that has been promised but not yet booked. The 200 million dollar slice is the price it put on that uncertainty.

Owners hear AI-native from every vendor pitching them this year, and most of those claims cannot be checked at signing. The structure Dassault used is the check. You do not have to argue about whether the AI works; you tie the money to the outcome and let the invoices settle it. That discipline is portable well beyond a billion-dollar acquisition, down to a renewal or a pilot contract.

What it changes for buyers and for the 200 customers

For the life-sciences firms already on ArisGlobal, the deal is a lock-in question before it is anything else. A standalone compliance tool is becoming a module inside a much larger platform sold by a company worth tens of billions, which usually means a broader roadmap and a stickier, pricier one. If you are among the 200, ask now what leaves your current contract untouched and what quietly migrates onto the wider stack over the next two renewals.

For everyone else, take the template and put it to work. When your next AI vendor quotes a price built on revenue it expects the AI to generate, propose the Dassault split: pay confidently for what exists, and make the rest wait on delivery. The deal closes in the second half of 2026 pending regulatory clearance, but the negotiating idea inside it is available to you on Monday.

Read next: Meta's Best New Model Is Closed, Paid, and US-Only  |  Two AI Giants Just Became Your Integrator