- The all-stock deal is valued at $4.5 billion, subject to adjustments, and is expected to close in Palo Alto's fiscal third quarter of 2027.
- Applied Compute says it passed $100 million in annualized revenue run rate in under 18 months, so the price is about 45 times that run rate and 1.4 times its November 2 valuation.
- Co-founder and CEO Yash Patil becomes Palo Alto's Co-Chief AI Officer, and the startup's research team, which came out of OpenAI's post-training group, joins the security company.
Palo Alto Networks has agreed to buy Applied Compute, an 18-month-old startup that helps companies retrain and run open-weight AI models on their own data, in an all-stock deal valued at $4.5 billion, 16 days after the startup raised money at a $3.25 billion valuation.
Palo Alto Networks will pay $4.5 billion in stock for Applied Compute
Applied Compute sells a platform for post-training and inference: it takes open-weight models, adapts them to a customer's data through fine-tuning and reinforcement learning, then hosts them in production. Yash Patil, Rhythm Garg and Linden Li founded it in 2025 after working on post-training at OpenAI. Palo Alto says it evaluated the company first as a prospective customer, and that its Unit 42 security team's Frontier AI Assessment of Applied Compute informed the decision. Palo Alto announced the deal alongside fiscal first-quarter results showing revenue up 39% to $3.44 billion and next-generation security annual recurring revenue up 66% to $9.73 billion.
A security company is buying a model-training team
Arora frames the purchase as a talent and defense move. In the release he argues that every enterprise will run a mix of open and closed models and that each needs securing, and on X he wrote that Patil's team must make Palo Alto the leader in applying frontier AI to its own security portfolio. Owning a training and serving stack lets Palo Alto build its own models and sell customers a way to run open-weight models inside public clouds or sovereign environments without sending data to a third-party lab. The release says Applied Compute will keep serving AI-native companies, digital natives and large enterprises, and that Palo Alto plans to bring the platform to its customer base of more than 70,000 organizations. Palo Alto also raised its fiscal 2027 outlook to revenue of $14.52 billion to $14.57 billion, so the acquisition lands while its core security business is still accelerating.
"Every enterprise is going to run a mix of open and closed models, and every one of those models needs to be secured."
Nikesh Arora, Chairman and CEO, Palo Alto Networks, acquisition release, November 18, 2026
The timing tracks the threat picture. The same week, European officials confirmed an AI-generated cyberattack on the European Commission linked to a framework that plugs commercial models into intrusion tooling, the kind of attacker capability Arora's industry now has to match.
Palo Alto paid a 38% premium over a valuation set just over two weeks earlier, for a company with roughly $100 million of run-rate revenue, and paid it in stock during a week when AI shares were falling hard. That combination shows how scarce proven post-training teams have become: a $9.73 billion recurring-revenue security business decided it was faster to buy one than to build one, and Applied Compute's investors accepted Palo Alto shares as the currency for that judgment.
Santage is committed to independent, transparent journalism. This article is produced in accordance with Santage's Editorial Standards and aims to provide accurate and timely information. Readers are encouraged to verify information independently.
--- Related: /news/2026/11/20/eu-commission-ai-cyberattack-artex, /news/2026/11/04/fal-16-billion-ai-inference-valuation, /news/2026/11/10/fireworks-ai-second-breach-model-weights Cover image: images/palo-alto-applied-compute-4-5-billion.jpg (1200x630), alt text from Shailey at conversion. Embeds: stamped valuation chart; Nikesh Arora tweet; primary-source quote; reader poll. Internal links: /glossary/open-weights, /glossary/fine-tuning, today's ANALYSIS.
Santage is committed to independent, transparent journalism. This article is produced in accordance with Santage's Editorial Standards and aims to provide accurate and timely information. Readers are encouraged to verify information independently.