Palo Alto Networks reaches agreement to acquire Applied Compute in a $4.5 billion all-stock transaction
Cybersecurity company Palo Alto Networks (NASDAQ: PANW) announced on November 18 that it has reached a definitive agreement with open-weight model post-training and inference platform Applied Compute to acquire the company in a $4.5 billion all-stock transaction, subject to adjustments. The company said the move aims to help enterprises securely train, adapt, and deploy open-weight models in public clouds and sovereign clouds.
Applied Compute co-founder and CEO Yash Patil will become co-chief AI officer at Palo Alto Networks after the transaction closes, and the Applied Compute team will also join the acquirer. The transaction remains subject to customary closing conditions, including regulatory approval, and is expected to close in Palo Alto Networks’ fiscal third quarter of 2027.
Palo Alto Networks Chairman and CEO Nikesh Arora said in a statement that every enterprise will run a combination of open and closed models, and every model will need to be protected. He added that Applied Compute has built the best post-training and inference platform they have seen, and its research team is also outstanding.
Applied Compute was founded in 2025 by former OpenAI researchers Yash Patil, Rhythm Garg, and Linden Li. The company had just completed a $350 million financing led by Elad Gil on November 2, with a post-money valuation of $3.25 billion. The acquisition announcement was released alongside Palo Alto Networks’ fiscal first-quarter results, and the company’s share price rose in after-hours trading.
Original source:
Title: Palo Alto Networks to Acquire Applied Compute, Bringing Secure Open-Weight AI to the Enterprise
Publisher: Palo Alto Networks
Publication date: 2026-11-18
URL: paloaltonetworks.com/company/press/2026/palo-alto-networks-to-acquire-applied-compute--bringing-secure-open-weight-ai-to-the-enterprise
Source type: official company press release
Verification status: The full text of the official press release was read and cross-checked against Reuters, Bloomberg, and other reports
Supporting content: $4.5 billion all-stock transaction (subject to adjustment); helping enterprises securely train, adapt, and deploy open-weight models in public clouds and sovereign clouds; Yash Patil to become co-chief AI officer; expected to close in fiscal Q3 2027, subject to regulatory approval.
ABAB AI Insight
This transaction reflects the path by which a cybersecurity giant is extending into enterprise AI infrastructure. Palo Alto Networks emphasizes a future of hybrid models in which enterprises will use both open-weight and closed frontier models, with security embedded throughout training, adaptation, and deployment. After acquiring Applied Compute, the company can directly offer customers the ability to post-train open-weight models on their own data rather than providing only perimeter security. This forms a continuation of its expansion of AI security capabilities following its 2026 acquisitions of CyberArk (identity security) and Chronosphere (observability).
Applied Compute reached an annualized recurring revenue run rate of $100 million in only about 18 months after its founding, and it completed financing at a $3.25 billion valuation in early November. The $4.5 billion purchase price represents an estimated premium of about 38% in just two weeks. The founding team came from OpenAI, and the platform supports enterprises in training and improving proprietary models on their own data, with customers including Microsoft and Shopify. Yash Patil also confirmed the agreement on X and stressed that customers will continue to own and control their own models.
The all-stock structure means Applied Compute shareholders will receive Palo Alto Networks shares rather than cash, causing dilution for existing shareholders while preserving the acquirer’s cash. The announcement came alongside stronger-than-expected fiscal first-quarter results, and the company also raised its full-year revenue outlook; the share price rose in after-hours trading. This shows the market’s recognition of the security plus AI model deployment combination and also reflects the current pace at which cybersecurity vendors are using acquisitions to quickly fill out enterprise AI deployment capabilities.
From a regulatory and closing perspective, the deal still requires customary approvals and is expected to close in fiscal third quarter of 2027, leaving timing and approval uncertainty. For Applied Compute, it is a path to a rapid exit from being an independent AI infrastructure company; for Palo Alto Networks, it is a key step in directly embedding its cybersecurity expertise into the lifecycle of enterprise open-weight models.
ABAB News · Cognitive Law
By acquiring an AI model post-training and inference platform, a cybersecurity company is extending its security capabilities from perimeter defense to the full process of model training and deployment.
An AI infrastructure company that is less than two years old is being acquired soon after completing high-valuation financing, reflecting the dual drivers of enterprise AI deployment demand and strategic buyers filling capability gaps.