Responsible AI

Grounded and cited

When KinetixAI answers from your sources, it cites them, so you can check the claim against the exact page, section, or line. We want answers you can verify, not answers you must take on faith.

Honest about limits

AI makes mistakes. We tell you when an answer may be uncertain, and we encourage you to confirm anything important against primary sources. KinetixAI is a learning aid, not an authority.

Focused scope

KinetixAI is built for AI, machine learning, systems, mathematics, and quantum topics. It is designed to stay on those subjects rather than act as a general purpose oracle.

Your data is not training data

We do not use the papers, notes, or questions in your workspace to train models. Your research is for your learning, not our model development.

Human oversight

People stay in control. You decide what to upload, which tier of explanation to read, and whether to act on an answer. We provide tools, not autonomous decisions about your work.

Fairness and access

We aim to explain the same concept at the level that fits each learner, from a plain-language story to the math and the code, so more people can reach the same understanding.

Reporting a concern

If an output seems harmful, biased, or wrong in a way that matters, tell us at hello@kinetixai.ai so we can look into it.

Evolving practice

Responsible AI is an ongoing effort. We will update these principles as the technology and our understanding improve.