Sovereign by Design
Episode 15 with Michael Burnett-Davies, the driving force behind Sovereign AI.
Michael Burnett-Davies is focused on a practical question many Australian businesses have not yet slowed down to ask: who actually owns the data, workflow and economics behind the AI tools they are starting to use?
In this episode of Queensland Business Stories, Michael joins Chris Tipper to talk about sovereign AI, local models, on-device deployment, data privacy, AI cost blowouts and the opportunity for small teams to build tools they can control.
AI That Stays In Your Environment
Michael’s definition of sovereign AI is practical. The model and the data stay inside the business environment, instead of being pushed into a centralised cloud platform where ownership, retention and future use can become murky.
That matters for any business, but especially for professional services, finance, legal and health teams handling sensitive client information.
The useful question is not just what AI can do. It is where the data goes while it does it.
The Hidden Cost Of Cheap AI
One of Michael’s strongest warnings is about economics. Many AI subscriptions feel affordable today because the providers are absorbing huge infrastructure costs. His concern is that once the pricing normalises, token-heavy usage could become far more expensive than business owners expect.
That is where local models become interesting. If a business can run the right model on existing hardware, or in a private environment, the long-term cost curve can look very different.
Integration Over Disruption
Michael is not arguing that every business needs a grand AI transformation project. In discovery work, he often finds that the useful answer is much simpler: map the existing workflow, automate the right steps, and use a smaller purpose-built model where it actually adds value.
That is a very different conversation from buying every expensive feature attached to the word “agent”.
Privacy Risks Are Already Here
The episode also gets specific about tools that send recordings, folders or work files to third-party AI providers. For some users that may be acceptable. For businesses with client confidentiality obligations, it can be a serious risk.
Michael’s point is not to avoid AI. It is to understand the terms, the data path and the alternatives before putting sensitive work through a tool built for convenience rather than control.
Small Teams Can Move Faster
Michael is optimistic about the opportunity for young people, small teams and owner-led businesses. Local AI does not remove the need for judgement, relationships or oversight. It gives people leverage when they understand the systems they are improving.
The businesses that learn early may not need to wait for a giant platform to package the solution for them.
Sustainability And Sovereignty
The conversation also touches on the physical footprint of AI: data centres, energy, water and the concentration of infrastructure. Michael sees local and private deployment as part of a broader discussion about sustainability and resilience, not only privacy.
Connect With Michael
Education and awareness: sovereignai.app
Managed private AI: sovereignai.group
More From Queensland Business Stories
Queensland Business Stories is presented by Connected Platforms. We sit down with Queensland business owners, operators and leaders to unpack the real decisions, constraints and lessons behind the work.