Enterprises want to run powerful AI inside their own cloud without sending private data to an outside API. Palo Alto Networks said revenue rose 39% to $3.44 billion in fiscal Q1 alongside a deal to buy Applied Compute to secure that setup.
Palo Alto Networks wants to own the risky middle between download and deployment.
Next-Generation Security ARR is annualized contract value from Palo Alto Networks cloud and AI security subscriptions. Palo Alto Networks said that metric hit $9.73 billion, up 66%. Nikesh Arora, chairman and chief executive of Palo Alto Networks, called the quarter a record Q1 start. Arora said customers turned to Palo Alto Networks to face an unprecedented threat environment as AI reshapes attacks.
Open-weight models are large models whose weights are published so enterprises can download and run them. Post-training is additional tuning that teaches a base model enterprise data and safety rules. Inference is live operation when a tuned model answers prompts inside enterprise systems. Applied Compute provides a platform for post-training and inference of open-weight models, Palo Alto Networks said. The idea is an enterprise pulls an open model, hardens it with Applied Compute tooling, then runs prompts through Palo Alto Networks policy checks.
The hard part is integration and trust. Palo Alto Networks warned the Applied Compute deal still needs closing conditions and Palo Alto Networks must merge Applied Compute operations without breaking deployments. Scale is real but costs show too, as Palo Alto Networks said GAAP operating income fell to $96 million from $309 million a year earlier.
Palo Alto Networks forecast fiscal Q2 revenue of $3.53 billion to $3.55 billion, Palo Alto Networks said. Palo Alto Networks pegged fiscal Q2 Next-Generation Security ARR at $10.20 billion to $10.22 billion. Palo Alto Networks raised full-year revenue guidance to $14.52 billion to $14.57 billion. Applied Compute brings research talent to support open-weight models, closed frontier models, or both. Watch whether Palo Alto Networks ships hardened model releases that stay patched after post-training.





