AI Compliance Checker
Who it's for: Businesses that already know they have privacy/compliance obligations and want to verify each AI tool in their stack actually meets them.
What it does: Select the AI tools you use and get a per-tool compliance checklist (confirmed, partial, or missing for each requirement), not just an overall score.
| Tool | DPA | Training | Residency | Staff trained | AI register | Privacy policy |
|---|
Recommended actions
Pre-filled data is based on publicly available vendor terms as of 1 July 2026. Verify current terms directly with each vendor before relying on this checklist. This tool provides general information only and does not constitute legal advice.
How this tool works
How the checklist works
The checker works tool by tool rather than calculating one blended score. First, select any of the 16 named AI tools your business uses, or choose “Other”. For every selected tool, the checker creates a separate checklist so that different products and plans are not combined into one result.
The first three items are pre-filled as Confirmed, Partially in place, or Not in place. They cover DPA status, model-training opt-out status, and data-residency status. You can edit every pre-fill to match your actual plan. Items 4 to 6 always begin as Not sure, because vendor terms cannot reveal how your business has configured or uses its account. The checker flags every item that is not fully confirmed and provides a relevant follow-up guide.
What the pre-filled answers are based on
The pre-filled answers are based on publicly available vendor terms reviewed as at August 2026. They are starting points, not findings about your particular subscription or settings. Vendor terms and product plans change, so compare each answer with current vendor documentation and the settings in your own account.
The August 2026 data includes partial Australian data-residency options for Microsoft 365 Copilot, Google Gemini Workspace, and ChatGPT Enterprise, Edu, or API. The checker records these as partial rather than treating them as universally available. Your plan still needs to be confirmed.
Worked example: ChatGPT Business and Zapier
Suppose a business selects ChatGPT Business and Zapier. The checker creates two independent checklists. For ChatGPT Business, it pre-fills the DPA item as Partially in place, training as Confirmed, and residency as Not in place. Zapier receives the same three pre-fills: DPA partial, training confirmed, and residency not in place. Items 4 to 6 start as Not sure for both products.
The confirmed training items are not flagged. The partial DPA items, residency items, and unanswered account-specific items are flagged because they are not fully confirmed. In particular, data residency appears as needing attention for both tools. The business can then verify each specific plan and edit the checklist if its current vendor documentation or account details support a different answer.
Frequently asked questions
Does the checker produce an overall compliance score?
No. It produces a separate checklist for each selected tool and flags items that are not fully confirmed.
Can I change a pre-filled answer?
Yes. Every vendor-based pre-fill is editable so you can reflect the plan your business actually uses.
Why do items 4 to 6 always say “Not sure”?
Those items depend on your business’s account settings and usage. The checker cannot determine them from public vendor terms.
What does a flagged item mean?
It means the item is not currently marked fully confirmed. Use the linked follow-up guide, current vendor documentation, and your account settings to investigate it.