🔍 Read the full analysis: Palo Alto Networks’ AI Expansion: Why It Wants To Run The Workloads on ThorstenMeyerAI.com
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Palo Alto Networks agreed on November 18, 2026, to acquire Applied Compute in an all-stock transaction valued at $4.5 billion, subject to adjustments and regulatory clearance. The deal would add tools for adapting and running open-weight AI models to Palo Alto’s security business, but the acquisition has not closed and its financial and integration details remain limited.
Palo Alto Networks has agreed to buy Applied Compute in an all-stock transaction valued at $4.5 billion, subject to adjustments, extending the cybersecurity company’s AI strategy into tools for training and running models. The companies announced the deal November 18; it remains subject to regulatory clearance and other closing conditions.
Applied Compute’s platform is designed for post-training and inference of open-weight AI models. In practical terms, that means businesses can adapt models whose weights are available using their own data, then run those models through the platform. Palo Alto says it plans to bring the service to its global customer base, while Applied Compute will continue serving AI-focused companies and large enterprises.
The transaction is all stock, rather than a cash purchase. Palo Alto’s announcement says the $4.5 billion valuation is subject to adjustments but does not describe them. It expects the acquisition to close in its fiscal third quarter of 2027, pending customary closing conditions, including regulatory clearance. Its fiscal first quarter ended October 31, 2026, placing the expected period roughly in February through April 2027.
Applied Compute chief executive Yash Patil is expected to become Palo Alto’s Co-Chief AI Officer after closing, and the company’s team is expected to join Palo Alto. The announcement does not name the other person, if any, who will share the role. Palo Alto announced the acquisition on the same day as quarterly results and said it would provide further transaction details on its earnings call.
Palo Alto Moves Beyond AI Security
The acquisition would expand Palo Alto’s role from securing AI systems and traffic into infrastructure used to build and operate enterprise AI. The company says businesses will use open-weight and closed models together and will need ways to adapt and run open models securely. This is Palo Alto’s stated view of the market.
Combining security products with model training and inference could allow customers to use fewer vendors across parts of an AI stack. It could also increase reliance on one provider for running AI workloads and protecting them. Businesses using either company’s products may seek clarity on data handling, service continuity, contract changes and exit options as the acquisition proceeds.
The transaction would add AI computing services to Palo Alto’s cybersecurity business. Chief executive Nikesh Arora told CNBC on the November 18 earnings call, “If it works, I think this will dwarf the ability that we have in cybersecurity.” He also said, “This could be bigger than what we do.” The comments describe Arora’s expectations and do not establish the deal’s future revenue or performance.
Products Worth Considering

Dell UltraSharp U4025QW 40-Inch Curved 5K2K WUHD LED Monitor

Apple Mac mini with M2 Chip, 16GB Unified Memory, 256GB SSD

Apple Mac mini Desktop Computer with M5 Pro Chip (2026)
As an affiliate, we earn on qualifying purchases.
A Third AI Deal in 2026
The Applied Compute agreement is Palo Alto’s third AI-related transaction announced in 2026, according to the supplied source material. Palo Alto completed its acquisition of AI gateway company Portkey on May 29 and announced a deal for AI agent platform Console on September 1. Together, the moves point toward a broader effort to cover several layers of enterprise AI, from agent traffic and security to the infrastructure on which models run.
Applied Compute describes its product as “the AI superCloud for open models.” In a November 2 company announcement, it said its platform, AC2, had supported more than 1,700 training runs since July and served more than 1 trillion tokens per day. It also reported growing from zero to $100 million in annualized revenue run rate in less than 18 months. These are company-reported figures; the supplied material says they have not been independently audited.
Applied Compute also said it raised $350 million in new funding at a $3.25 billion valuation in that November announcement. Its reported valuation before the Palo Alto deal and the acquisition’s $4.5 billion stated value are different figures, and the deal value is subject to adjustments. The announcement does not provide a breakdown of the transaction terms.
““If it works, I think this will dwarf the ability that we have in cybersecurity.””
— Nikesh Arora, Palo Alto Networks chief executive, quoted by CNBC
Deal Terms and Integration Remain Open
The acquisition is not yet complete. Regulatory clearance and other customary closing conditions remain, and the companies have not provided a detailed account of the adjustments that could affect the $4.5 billion transaction value. The supplied source material also says it could not access Palo Alto’s filings with the U.S. Securities and Exchange Commission or its investor-relations page, so the account relies on company announcements, Applied Compute’s website and attributed media reporting.
It is not yet clear how Applied Compute’s services will be integrated with Palo Alto’s products, what customers’ pricing or service terms may become, or how the combined company will handle customer data and model workloads. The announcement says Applied Compute will continue to serve its existing customer groups, but does not set out detailed post-acquisition commitments or a transition plan.
Applied Compute’s operating figures, customer examples and performance claims are its own reports. The supplied material describes no independent audit of those figures. Palo Alto’s claims about the need for secure open-model services and the possible scale of the market are strategic judgments, not confirmed measurements of future demand or financial results.
Regulatory Review and Closing Timeline
Palo Alto expects the transaction to close in its fiscal third quarter of 2027, a period that runs roughly from February through April 2027 based on its fiscal calendar. That schedule depends on regulatory clearance and the other closing conditions; the companies have not reported that those steps are complete.
Palo Alto said it would provide further details on the transaction during its earnings call. Readers and customers can watch for additional disclosures about the deal’s adjustments, integration, leadership structure and how Applied Compute’s platform will be offered. Until the acquisition closes, the announced plans—including Patil’s expected role and the transfer of the Applied Compute team—remain subject to completion of the deal.
Key Questions
What did Palo Alto Networks agree to buy?
It agreed to acquire Applied Compute, a company whose platform helps businesses adapt and run open-weight AI models.
How much is the acquisition worth?
Palo Alto says the all-stock transaction is valued at $4.5 billion, subject to adjustments. The announcement does not specify those adjustments.
Has the acquisition closed?
No. The deal remains subject to regulatory clearance and other customary closing conditions. Palo Alto expects it to close in its fiscal third quarter of 2027.
What will Applied Compute’s chief executive do?
Applied Compute CEO Yash Patil is expected to become Palo Alto’s Co-Chief AI Officer after closing. The announcement does not identify who, if anyone, will share the role.
Why is a cybersecurity company buying an AI platform?
Palo Alto says businesses will use open and closed AI models and need secure ways to train, adapt and run them. That is the company’s stated market rationale; the deal’s future business impact is not yet known.
Source: ThorstenMeyerAI.com
