Palo Alto Networks buys Applied Compute for $4.5bn to expand AI security reach
Palo Alto Networks will acquire Applied Compute, an AI startup, in an all-stock deal valued at $4.5bn. Applied Compute, founded by former OpenAI researchers, lets firms train and run open-weight AI models. The deal, announced alongside Palo Alto’s latest results, is expected to close in its fiscal third quarter of 2027, pending regulatory approval. Applied Compute’s CEO, Yash Patil, will become Palo Alto’s co-chief AI officer.
Palo Alto said it plans to let companies train, adapt and run open-weight models securely, a service it claims no large provider offers today. Applied Compute will continue to serve its own customers, while Palo Alto will offer the platform to its client base. The company expects the deal to add over $1bn in annual recurring revenue by fiscal 2029 and will invest in more than 100 megawatts of inference capacity, keeping capital spending under $3bn through 2030.
For regulated industries, the acquisition signals a shift: Palo Alto is moving beyond cybersecurity into broader AI infrastructure. The focus on open-weight models may appeal to banks, insurers and healthcare groups seeking more control and security over AI systems. The deal’s outcome depends on regulatory clearance and integration of Applied Compute’s team and technology.
Palo Alto’s raised forecasts and strong quarterly results suggest confidence in its expanded strategy. The company links recent demand to AI-driven IT risks, pointing to ongoing security challenges for enterprise users.
Find out more here