What is sent for an answer
When you submit a question to Ask LLM.how, your question and answer-style selection are sent to our server. Generated answers use the Mojica AI Gateway with selected public catalog records and source links. Processing uses our own AI infrastructure and may use Microsoft Azure OpenAI as a permitted fallback. Some reviewed reference answers and calculated comparisons do not require an AI request.
No model training
Your submissions and AI responses are not used to train AI models. Using content to answer your request or investigate a problem is separate from training or fine-tuning a model.
Diagnostic retention and access
We may retain submitted text, supplied request context, and generated responses for up to 30 days for troubleshooting, security, and abuse prevention. Relevant records may be retained longer when reasonably necessary to investigate a security incident or suspected misuse, resolve a dispute, or comply with legal obligations. Access is restricted to authorized personnel. Extended retention is periodically reviewed, and records are deleted when no longer needed.
Operational records may include a request identifier, selected tool, processing route, timing, outcome, and usage counts. Do not submit passwords, API keys, private documents, personal data, or confidential information.
Processing services and separate copies
Our diagnostic retention period is not a zero-retention promise for every processing service. Microsoft may perform abuse monitoring and retain flagged content under its own policies. See Microsoft’s data, privacy, and security information. Copies you save, download, or send to another service are separate from gateway diagnostic records.
Hosting, measurement, and outside links
Hosting, security, and site-measurement services may process ordinary request and usage information. This notice does not authorize sending your question or generated answer to traffic analytics. Model platforms and other sites linked from LLM.how have their own privacy policies.
Privacy questions
Contact the publisher with the feature, approximate request time, and request identifier if available. Do not include credentials or sensitive conversation content in a general contact message. About LLM.how and its publisher.