AIQ AI and Data Privacy

We understand that you may have questions about

AIQ AI and theData Privacy

Use this page to explain privacy ofexpectations yourfor data.

AI-assisted

WillAIQ Appvance’sworkflows, including AISG and other AI usefeatures. yourKeep customer data tohandling trainaligned with contractual, security, and regulatory requirements.

Privacy guidance

• Do not include passwords, API keys, tokens, personal data, regulated data, or customer-sensitive records in prompts unless explicitly approved.
• Use synthetic, masked, or minimized data for examples and generated tests whenever possible.
• Document what data is used by AI-assisted workflows and who approved that usage.
• Review generated output before saving it into shared documentation, tests, or repositories.
• Follow customer retention, access-control, and audit requirements for generated assets.

Operational checklist

1. Confirm the AI?

No. Our AI modelsfeature, wereinput genericallydata, trainedand onexpected whatoutput theybefore need.use.

Remove Theyunnecessary willsensitive not learndetails from yourprompts specificand dataexamples.Validate generated test logic before execution.Escalate unclear data-handling questions to the customer owner or applicationsecurity flows except for use with your own application under test in an isolated environment. In addition, AIQ does not store, or use ANY Personally Identifiable Information (PII) data.

What AI models are used in AIQ and how were they trained?

team.

Many AI and ML models are used in AIQ. These models are based on years of learning about applications in general, including actions within UX libraries. Any tuning by the user is only used for creating scripts for that specific application and would be unusable for any other purpose.

Examples:

Self-healing is a type of machine learning that learns each time a script is run and builds additional data around accessors.

See Script Optimization and Self-Healing for an example of self-healing in web testing.

AI Script Generation (AISG) is a form of generative AI/AI code generation not based on a transformer model, but on some twenty underlying models that together generate actual test scripts. AISG uses the application under test as a user would and creates thousands of scripts autonomously after a single training session with a QA engineer. The output of the AISG process are scripts, an application map, results, and a dynamically-built Blueprint file that is only applicable to that application.

See AI Script Generation 3.0 in AIQ for more information.

GENI is an upcoming technology which uses several AI models including a transformer, LLM, multimodal and others. All models are private, fixed, and do not learn from your application or data.