Introduction

In our first article, we explored the concept of the AI PC and what it means when language models move onto the device itself. 

Now we take the next step.

Because if AI is a new runtime, where should it run? In the cloud, on the device, or both? Which approach makes the most sense for you, your users, and your data?  Most people are now familiar with cloud-based solutions like ChatGPT, Microsoft Copilot, or Apple Intelligence. What far fewer people realize is that it is now possible to run advanced AI models locally — without subscriptions, without logging, and without the cloud.

What is Foundry Local?

Foundry Local is Microsoft’s solution for running AI models locally in secure and controlled environments. It is designed for organizations operating in closed or air-gapped networks, where cloud access is either not possible or not desired.

What is LM Studio?

LM Studio is an open-source–based tool that makes it easy to download, run, and interact with large language models locally on your Windows, Mac, or Linux machine. It provides a graphical interface where you can choose between different open-source models — entirely without the need for an internet connection.

Why does this matter?

Because the use of AI is no longer a simple “yes or no” decision. It is an architectural choice:

• Where is data processed?
• Who has access?
• How do we protect against unwanted behavior, data leakage, or compliance violations?

With cloud AI, you get power and scalability. With local AI, you gain control and data sovereignty. The two can be combined — but doing so requires a conscious and deliberate choice.

Examples of local AI usage

• Legal departments working with confidential documents
• Developers who want AI-powered code suggestions without sending IP to the cloud
• Scenarios where systems are generally offline but still require text summarization or data processing
• Organizations operating air-gapped networks

This is not about choosing sides. It is about understanding your options.

If you only know cloud AI, you risk overlooking a solution that could be more secure, more flexible, and better suited to your specific needs.