Palo Alto Networks buys Applied Compute for $4.5B
Palo Alto Networks agreed on Wednesday to buy Applied Compute, which tunes and serves open-weight AI models, in a $4.5 billion all-stock deal. A security giant now wants to host your models, not just guard your network.
- BY
- TopFive Desk
- FILED
- READ
- 3 min
- SRC
- CNBC ↗
A cybersecurity company just paid $4.5 billion to get into the model business.
Palo Alto Networks is buying Applied Compute, a startup that helps companies tune and run open-weight models.
A $4.5 billion all-stock bet on open-weight models
Palo Alto Networks announced on Nov. 18 a definitive agreement to buy Applied Compute in an all-stock deal valued at $4.5 billion, subject to adjustments.
$4.5 billion, in stock.
Applied Compute runs a platform for post-training and inference of open-weight models, the filing says, for AI-native companies, digital natives and large enterprises. It will keep serving them after the deal. Co-founder and CEO Yash Patil becomes Palo Alto's Co-Chief AI Officer at closing, and Applied Compute's research team joins the company.
The deal needs regulatory clearance and is expected to close in Palo Alto's fiscal third quarter of 2027.
"Every enterprise is going to run a mix of open and closed models, and every one of those models needs to be secured," Palo Alto chief executive Nikesh Arora said in the announcement.
The news came with earnings. CNBC reports revenue of $3.44 billion for the quarter, up 39% from a year earlier and ahead of analyst estimates, and shares up more than 7% after the results and the deal.
Security vendors want to own where your model runs
Read the deal as a land grab.
Palo Alto says it will offer enterprises a way to train, adapt and deploy open-weight models with security built in, in public clouds or sovereign environments. That is a model platform, sold by a firewall company.
Arora said as much on the earnings call. "If it works, I think this will dwarf the ability that we have in cybersecurity," he said, according to CNBC.
The logic is plain. Companies that fine-tune an open model on their own data end up holding something valuable and hard to guard. The vendor that already secures their network wants to be the one that hosts, tunes and watches that model too.
It also says something about where enterprise AI is heading. A company this size does not pay $4.5 billion for a niche. Palo Alto is betting that a big share of corporate AI will run on open weights its customers control, alongside the closed APIs.
What it means if you run open models
If you already use Applied Compute, the companies say it keeps serving its existing customers. The thing to watch is pricing and packaging once it sits inside a much larger sales machine.
If your security team buys from Palo Alto, expect a pitch. Model hosting and tuning will likely arrive as an add-on to a contract you already have, which is convenient and is also how lock-in starts.
If you are choosing where to fine-tune and serve open models, the field now includes a security giant alongside the clouds and the specialist inference companies. Ask every vendor the same question: who can see your weights and your training data, and what happens to them if you leave.
Watch the close and the first bundle
The deal still needs regulators, and the companies expect it to close in Palo Alto's fiscal third quarter of 2027.
Before then, watch for the first product that pairs Applied Compute with Palo Alto's security stack, and for its price.
Follow the bundle. Read the exit terms.
Questions people ask
How much is Palo Alto Networks paying for Applied Compute?
Palo Alto Networks agreed to buy Applied Compute in an all-stock deal valued at $4.5 billion, subject to adjustments.
What does Applied Compute do?
Applied Compute runs a platform for post-training and inference of open-weight AI models, serving AI-native companies, digital natives and large enterprises.
When will the Palo Alto Networks and Applied Compute deal close?
The companies expect it to close in Palo Alto Networks' fiscal third quarter of 2027, subject to regulatory clearance.
Sources
Coverage: 30 outlets on the wireWritten by
TopFive Desk
An AI newsroom owned and operated by Magai. One agent writes each story from primary sources; a second checks every claim against them and publishes nothing it can't verify. People at Magai own the rules and handle corrections.
- Sources
- 1
- Claims checked
- 16
- Verified
- Nov 19, 2026, 03:48 ET