The Daily Ledger

AINews

Palo Alto Networks turns to open-weight AI with a $4.5 billion acquisition

Palo Alto Networks has agreed to buy Applied Compute in an all-stock deal valued at $4.5 billion, aiming to add tools for securely training and deploying open-weight AI models.

The Daily Ledger Staff · Updated 4 min read
A technology professional monitors a glowing neural network connected to modular AI components, secure shields and enterprise server racks.
Illustration generated with AI

Palo Alto Networks has agreed to acquire Applied Compute in an all-stock transaction valued at $4.5 billion, adding a specialist platform for training, adapting and running open-weight artificial intelligence models. The cybersecurity company announced the agreement on November 18, saying the combination is intended to help businesses use models they can customise while keeping security controls in place.

Applied Compute develops technology for post-training and inference. Post-training is the work of adapting a model after its initial training so it can perform more effectively for a particular task or organisation. Inference is the process of using a trained model to generate an answer or carry out a task. The two capabilities are important for companies that want to run models on their own data rather than rely only on a general-purpose service.

Why open-weight models are attracting enterprise interest

Open-weight models make their trained parameters available for organisations to download, adapt or deploy under the terms of their licences. That can give businesses more control over where a model runs and how it is adjusted. It does not automatically mean that every part of a model's training data or development process is open, and the legal terms can differ between models.

For companies operating in regulated sectors or across multiple countries, the choice of deployment environment can matter as much as model performance. Some organisations want to keep sensitive information inside their own infrastructure or use a sovereign cloud environment that meets local requirements. Others need to connect models to internal databases and software while limiting what an automated system can access.

Palo Alto Networks said Applied Compute's platform will be used to help enterprises train, adapt and deploy open-weight models in public-cloud and sovereign environments. The company also expects the acquired research team to contribute to AI development across its wider product portfolio.

What the acquisition brings

Applied Compute will continue serving AI-native companies, digital businesses and large enterprises, according to the announcement. Its team is expected to join Palo Alto Networks after the transaction closes. Co-founder and chief executive Yash Patil is expected to become Palo Alto Networks' co-chief AI officer following completion.

The proposed deal is valued at $4.5 billion, subject to adjustments, and is structured as an all-stock transaction. It remains subject to customary closing conditions, including regulatory clearance. Palo Alto Networks expects the acquisition to close in its fiscal third quarter of 2027, but that timetable is a company expectation rather than a completed transaction.

The agreement is part of a wider strategic shift among technology and cybersecurity companies as enterprise use of AI expands. Security vendors are seeking to protect not only the data and infrastructure used by AI systems, but also the models, tools and automated agents that interact with business applications. At the same time, companies building AI products are looking for ways to demonstrate that their systems can be operated within an organisation's existing security framework.

Security will be central to adoption

Running a model inside a company's environment does not remove security risks. Organisations still need to manage access permissions, protect training and inference data, monitor model behaviour and understand how connected tools can be used. A model that can retrieve information from internal systems may be useful, but it also needs clear limits on which users and processes can reach that information.

Open-weight systems can give customers flexibility, but that flexibility also places more responsibility on the organisation deploying them. Businesses must decide how to test model updates, record changes, assess third-party components and respond when an application behaves unexpectedly. Security controls must cover the model lifecycle rather than only the network on which it runs.

What remains uncertain

The acquisition announcement does not establish that every planned integration is already available, nor does it guarantee that customers will see immediate product changes. The transaction must first close, and the companies will then need to combine teams, technology and customer support. Pricing, product packaging and the timing of specific new services have not been detailed.

The strategic bet is that enterprises will use a mix of closed frontier models and open-weight systems, and will want a consistent way to secure both. Whether the deal delivers that result will depend on how well Applied Compute's capabilities fit into Palo Alto Networks' existing products and how customers respond once the combined offering reaches the market.

Sources

This story was prepared with the help of automated drafting tools and was reviewed and fact-checked by our editors before publication. See our Editorial Policy.

Report a problem with this story · Copyright & takedown