AI Trust Center
AI trust
Responsible AI at Logiqc
Learn how Logiqc's AI features handle your data, which data is sent for processing, and the safeguards in place. Logiqc's AI features use the Azure OpenAI Service from Microsoft, and only the data needed to generate your requested response is sent for processing.
- Not used to train AI models
- Prompts discarded after use
- Stored data stays in Australia
Reviewing Logiqc AI?
Our team can answer AI questions as part of your security or privacy review.
Ask about AI View subprocessorsPrivacy and protection for your data
These commitments apply every time you use an AI feature in Logiqc.
No model training
Customer data submitted through Azure OpenAI is not used to train, retrain, or improve Microsoft's foundational models or any third-party models.
Transient processing
Data is processed in memory only to generate the response, and is discarded as soon as the request is complete, unless it has been flagged by Azure for abuse monitoring.
Stored data stays in Australia
AI outputs and system records stored at rest remain in Microsoft Azure data centres in Australia.
Only what is needed
Only the data needed to generate the requested response is sent for processing.
Attachments are not sent
Files uploaded or attached to an incident report are not submitted for AI processing, unless submitted by a user or required for the task.
TODO - Network Isolation swap out. Enterprise security standards
Data is processed under Azure OpenAI enterprise security standards, and access to temporary monitoring logs is strictly controlled by Microsoft.
How AI requests are processed
Each AI request follows the same path, from Logiqc to the Azure OpenAI Service and back.
Step 1Australia
You use an AI feature in Logiqc
Logiqc gathers only the content needed to answer your request.
- Form content and inputs
- Context the prompt needs
- Attachments are left out
Step 2Scoped region
Azure OpenAI Service processes the request
The deployed AI models generates a response in memory.
- Not used for model training
- Discarded when the request completes
- Azure abuse monitoring applies
Step 3Australia
The response returns to Logiqc
You review the response before using it.
- Saved outputs stored in Australia
- AI usage recorded for audit
What data is sent
When you use an AI feature, only the data strictly necessary to generate the requested response is sent to the Azure OpenAI Service.
- Form context: text entered into fields, selected dropdown values, category selections, and configuration settings such as Severity settings.
- Contextual data: any operational or background context the prompt needs to return an accurate result.
- Dynamic select field content: supporting select fields with dynamic content are sent as readable text.
- Prompts after use: prompts are discarded once the response has been generated.
- Training data: your data is not used to train or improve AI models.
Prompts are not de-identified
Prompts are not de-identified, pseudonymised, masked, or redacted before they are sent for processing. Keep this in mind when deciding what to enter into fields that use AI.
Audit history and traceability
Logiqc keeps a record of AI use so you can see how AI features are being used across your organisation.
- Who used the AI
- How many tokens and credits were used
- When the AI was used, and on what type of activity in the platform
- Which incidents were submitted with generative AI
- "A/B" records showing which suggestions the AI made and how the user edited them
Azure abuse monitoring and review
Azure abuse monitoring is designed to detect and reduce harmful or non-compliant use of the service. It uses content classification, pattern detection, and review processes to identify potential breaches of Microsoft's Code of Conduct.
Automated review
By default, prompts and completions flagged as potentially harmful are reviewed using automated AI-based analysis.
Human review only when needed
Human review is not automatically enabled. If automated checks do not meet confidence thresholds or are unavailable, authorised Microsoft employees may review flagged content under strict security controls.
Monitoring stays on
Microsoft only allows Azure abuse monitoring to be disabled under special consideration. Logiqc has not sought this permission.
AI FAQs
Answers to the questions we are asked most often about AI.
Is our data used to train AI or fine tune models?
No. Customer data submitted through Azure OpenAI is not used to train, retrain, fine-tune or improve Microsoft's foundational models or any third-party models.
Where is our data processed when we use AI? What do you mean by "Scoped Region"?
Inference, the step where the AI model processes your prompt, is not able to be restricted to a single country at the moment. Logiqc's policy is to always select the smallest available region containing Australia.
AI outputs and system records stored at rest remain in Microsoft Azure data centres in Australia.
Which AI model does Logiqc use?
Logiqc currently uses models fit for the specific task it is performing. We use the Microsoft Azure platform for this where only models that have been vetted by Microsoft are available for use.
Are prompts de-identified before they are sent?
No. Prompts are not de-identified, this information is required to successfully complete the tasks asked of the AI. Supporting select fields with dynamic content are also sent as readable text. Prompts are discarded after the response is generated.
Can anyone at Microsoft see our prompts?
Only in limited cases, where the Abuse monitoring is triggered, detailed above.
What does Logiqc record about AI use?
Logiqc records who used the AI, how many tokens and credits were used, when, and on what type of activity in the platform. Logiqc also records which incidents were submitted with generative AI. Logiqc does not keep records of the AI's suggestions and how users edited them.
Can we turn AI features off?
Our AI integrations are opt-in, putting the power of choice with you.