Businesses can control costs, protect sensitive data and reduce their reliance on the cloud by running AI models locally, Brilyant CEO Akash Saxenaa said at TechSparks 2026.
Saxenaa's argument starts from how far AI tools have come. They have made it easier for companies to analyse documents, generate content and automate routine tasks. That ease of use has pushed adoption across many kinds of businesses.
On-device AI is the alternative he pointed to. Instead of sending every request to a remote cloud service, the model runs on hardware the business controls, such as a laptop, a phone or its own servers. The pitch rests on three things: spending that is easier to manage, tighter handling of sensitive information, and less exposure to an outside provider.
The data question matters most for companies that work with confidential documents. When a model runs locally, the material it processes does not need to leave the organisation's own environment. That is the core of the privacy case for the approach.
Cost is the other driver. Cloud-based AI services are typically used on an ongoing basis, so spending can grow with usage. Moving some workloads onto local models is one way for companies to gain more control over that spending.
Saxenaa's remarks reflect a wider debate among businesses about where AI workloads should run, and how much of that work should depend on a third-party cloud provider.
