Healthcare
Healthcare organisation, Australia
Customer story
Australian Healthcare Organisation
How an Australian healthcare organisation used platform controls and pre-submission guardrails to reduce Shadow AI prompts by 92.8% in 30 days.
Read the full storyKey outcome
92.8%
fewer Shadow AI prompts in 30 days
Measured after Aona’s guardrails were introduced
Aona gave us visibility into which AI platforms were being accessed across the organisation and helped us proactively discourage use of unapproved tools while reinforcing Copilot as our approved option. It has been easy to deploy, lightweight for end users, and a valuable addition to our AI policy.
- Shadow AI prompts after guardrails
- 446 → 32
- unapproved AI platforms contained
- 7 → 2
- raw prompt or file contents retained
- 0
Customer context and approach
The challenge
An Australian healthcare organisation had approved Microsoft Copilot, but staff were also using consumer AI tools to draft emails, summarise documents and search. Aona surfaced 446 Shadow AI prompts submitted outside the approved tool, alongside activity on more than seven unapproved AI platforms.
The approach
The organisation deployed Aona's browser extension and governance portal to detect AI-site visits and intercept prompts and file uploads before submission. It blocked unwanted platforms and subscriptions, applied real-time guardrails to authorised platforms and guided staff back to Microsoft Copilot. Prompt and file contents were processed transiently in Australia and not retained; only high-level analytics were stored in Australia.
- Shadow AI Detection
- Browser Extension
- Real-time Guardrails
- Healthcare
