BoardPro AI is the set of AI capabilities built into BoardPro. It is not a separate product, and it does not send your board information to a consumer AI tool.
This article covers how BoardPro AI handles your information. For BoardPro's general security – hosting, encryption, multi-factor authentication and access levels – see BoardPro Security.
Last reviewed: 26 August 2026. Subject to change.
What BoardPro AI can access
BoardPro AI has four capabilities. Two work with content already in your BoardPro account:
AI Assistant – answers questions across your board packs, minutes and decisions
AI Agenda Builder – drafts an agenda structure from a description you provide
Two work with a file you choose to upload:
AI Minutes – drafts minutes from a meeting transcript you upload
Annual Work Plan import – converts an existing work plan document into a governance calendar
No capability goes looking for content on its own. Uploads are a deliberate action taken by a signed-in user.
BoardPro does not record your meetings
BoardPro has no recording capability. It does not join, listen to, or record meetings, and it does not accept audio or video files.
For AI Minutes, you record the meeting using your own conferencing tool – Zoom or Microsoft Teams, for example – and upload the transcript file afterwards.
Permissions
BoardPro AI does not have its own access level or service account.
AI Assistant can only reach documents the signed-in user is already permitted to open. If a user cannot open a document in BoardPro, the AI cannot read it on their behalf. Your existing access levels are unchanged by AI.
Where AI processing happens
AI processing is performed by Microsoft Azure OpenAI in the Australia East region – the same Microsoft region that hosts BoardPro.
Your board information is not sent to public or consumer AI tools such as ChatGPT.
The model, region and retention details for each AI capability are published on our AI Functionality Providers page, which is updated when they change. That page is the current source for those details.
Your content is not used to train AI models
Your board information is never used to train, retrain or fine-tune any AI model. This is a contractual term with our AI provider, not only a policy position.
What happens to an uploaded transcript
The transcript is used to generate your minutes and is then deleted from BoardPro storage. Microsoft retains the request in the Australia East region for up to 30 days for abuse monitoring only. It is not used for training.
Separation from other customers
BoardPro is a multi-tenant platform, and customer data is logically separated. AI processing is scoped to your organisation's content, and to the access level of the person making the request.
AI output is always a draft
BoardPro AI does not publish anything.
Minutes, agendas and answers are produced as drafts inside BoardPro for a person to review, edit and approve. Nothing enters your board record without that step, and accountability for the record stays with the board rather than with the AI.
Alongside this, Microsoft Azure OpenAI applies content safety controls to AI requests.
Can we turn BoardPro AI off?
Yes. AI features are controlled at organisation level. Administrators can enable or disable them for the organisation.
Can someone influence AI answers by uploading a misleading document?
BoardPro AI answers from the documents held in your BoardPro account. It does not assess whether a document is authentic or accurate. If a misleading document is in your repository, the AI may reflect what it says.
Three things control this:
Who can upload. Only Administrators, Board Secretaries, Chairs and Senior Executives can add documents to Governance Documents. Board Members and Executive/Guest users cannot.
Sources are shown. AI Assistant shows the documents an answer draws on, so you can see where a statement came from and check it against the source rather than taking the answer on trust.
Permissions still apply. The AI only reads documents the person asking is already permitted to open, so a document cannot reach anyone who could not already see it.
Can the AI be made to take actions it shouldn't?
BoardPro AI produces text. It cannot take actions in BoardPro, change permissions, send messages, or reach systems outside your BoardPro environment. Any output appears as an editable draft that a person reviews before it becomes part of the board record.
Security reviews and questionnaires
If your IT or security team is reviewing BoardPro, the quickest starting point is our listing in the Cloud Security Alliance STAR Registry.
The listing contains our completed CAIQ – a standardised set of more than 250 security control questions covering areas such as data security, access control, infrastructure and incident response. It's a self-assessment published publicly, in a format your team can compare directly against other providers.
View it here: Cloud Security Alliance STAR Registry – BoardPro
Most security reviews are answered in full by that document. If something isn't covered – including questions specific to BoardPro AI – contact us through the Support button in BoardPro or at support@boardpro.com.
Disclaimer: This information was last reviewed on 26 August 2026 and is subject to change.
Related articles