Palo Alto Networks, an AI cybersecurity company, has agreed to buy Applied Compute for $4.5 billion in stock. The deal would give its customers a way to adapt open-weight AI models using their own data and run them with security built in.
Applied Compute, a platform for post-training and inference of open-weight AI models, helps businesses fine-tune an existing model for a specific task and then use it to generate answers or predictions. “Open-weight” means the model’s learned settings are available for others to download and use.
“Applied Compute has built the best post-training and inference platform we’ve seen, and its research team is exceptional,” said Nikesh Arora, chair and CEO, Palo Alto Networks. “We saw that firsthand as a prospective customer, and our Unit 42 Frontier AI Assessment confirmed their strong security foundations.”
Palo Alto Networks expects businesses to use a mix of open-weight models they can adapt themselves and closed models that providers control. It plans to offer Applied Compute’s platform to customers worldwide so they can train and run open-weight models in the public cloud or in systems set up to meet local rules on data. Applied Compute will continue to serve companies built around AI, digital businesses, and large enterprises.
Applied Compute’s research team will also help train Palo Alto Networks’ own large and small AI models for use across its products. Yash Patil, Applied Compute’s co-founder and CEO, will become Palo Alto Networks’ co-chief AI Officer after the deal closes.
“Joining Palo Alto Networks will put security at the center of all our products and accelerate our traction with the world’s largest organizations,” said Patil. “I’m excited for our team to continue its mission alongside our talented new colleagues as part of Palo Alto Networks.”
Palo Alto Networks expects the deal to close in its fiscal third quarter of 2027, subject to regulatory clearance and other customary closing conditions. The $4.5 billion value is subject to adjustments.