Palo Alto Networks agrees to buy AI startup Applied Compute for $4.5 billion in stock

Startups & Funding

Palo Alto Networks Pays $4.5 Billion in Stock for AI Startup Applied Compute

Palo Alto Networks will pay $4.5 billion in stock for open-weight AI startup Applied Compute — 38% above its Nov. 3 valuation — with CEO Yash Patil becoming Co-Chief AI Officer.

By
Rebecca Stone
Filed
Channel
Startups & Funding
Read
3 min read

Palo Alto Networks has agreed to pay $4.5 billion in stock — about 13.3 million new shares — for Applied Compute, a startup that helps enterprises customize and run open-weight AI models. The deal, announced Wednesday, Nov. 18, works out to roughly 1.6% dilution of the cybersecurity company's 821 million outstanding shares, by MLQ.ai's calculation based on the company's 8-K filing.

The price marks a sharp premium. Applied Compute was valued at $3.25 billion in a $350 million round reported by SiliconANGLE on Nov. 3, meaning Palo Alto is paying about 38% more, less than three weeks later.

What are the terms?

Palo Alto will issue common stock at $337.16 a share, the volume-weighted average closing price over the 15 trading days through Nov. 17, according to the 8-K filing. The final share count is not yet known, the filing says. Closing depends on customary conditions, including the end of the Hart-Scott-Rodino antitrust waiting period, and is expected in Palo Alto's fiscal third quarter of 2027.

Applied Compute co-founder and CEO Yash Patil will become Palo Alto's Co-Chief AI Officer once the deal closes, the company said.

The announcement landed alongside Palo Alto's fiscal first-quarter results, in which it raised its full-year outlook.

Why does a security company want an AI startup?

Open-weight models are AI models whose trained parameters are published, so a company can adapt them to its own data and run them on infrastructure it controls. Applied Compute sells the tooling for exactly that work — post-training, the tuning of a model after its initial training, and inference, the running of a deployed model.

Palo Alto's argument is simple. Enterprises will run a mix of open and closed models, and a large share will pick open-weight models tuned to their own data, cost and control requirements. It says no large provider currently delivers those models securely, and that it intends to offer them in public clouds and sovereign environments. The Applied Compute research team will also help train Palo Alto's own AI models.

"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 Chairman and CEO Nikesh Arora said.

In other words, the deal is a bet that security becomes the gating factor for enterprise AI adoption — and that Palo Alto can own that gate.

How fast did Applied Compute rise?

Fast. Patil founded the company with Rhythm Garg and Linden Li, all former OpenAI researchers.

The funding trail tells the story:

  • April: $80 million raised at a $1.3 billion post-money valuation, in a round led by Kleiner Perkins, according to the company's law firm, Latham & Watkins.
  • Nov. 3: $350 million raised at a $3.25 billion valuation, led by Elad Gil, with Nvidia, Shopify Ventures and HubSpot Ventures participating, SiliconANGLE reported.
  • Nov. 18: agreed acquisition at $4.5 billion.

That is a roughly 3.5x valuation climb in about seven months. SiliconANGLE also reported that Applied Compute's annualized recurring revenue had passed $100 million, and that its customers include Microsoft and DoorDash.

For Palo Alto, the question now is execution: turning a $100 million-ARR startup into the secure delivery layer for open-weight models across public and sovereign clouds, before rivals stitch together the same capability.

Source: mlq.ai

Share this article:

More from Rebecca Stone

Rebecca Stone

Show full bio

Correspondent covering technology and product news in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,131 articles

Related articles

« Previous article