These Fair Use of AI terms (“AI Terms”) govern your use of AI-powered features across the QAI platform. They are a separate agreement from, and in addition to, the QAI Terms of Service and Privacy Policy. The general principles below apply to every AI feature; feature-specific terms (such as QVC AI Feedback) follow in the clauses further down. By using an AI feature — or, where a feature asks for it, by checking its consent box — you agree to these AI Terms.

1. Our Responsible AI commitment

Every AI feature QAI ships is held to these commitments:

• Transparency. We tell you when AI is used, what it does, and which provider processes your data.
• Human oversight. No decision with a legal or similarly significant effect on you is ever made solely by AI; a person is in the loop.
• Data minimization. Only the data needed for the feature reaches the model; sensitive data is withheld by design.
• No training on your data. We do not use your inputs to train AI models, and we send them to providers under terms that do not permit training on your data.
• Accuracy & honesty. We present AI output as a fallible aid, not ground truth.
• Security. Requests are encrypted in transit and scoped to your own data.

2. Fair use of AI features

AI features are provided for your own legitimate use. You agree not to:

• Generate unlawful, deceptive, infringing, or harmful content (the AI-use standards in the Terms of Service apply).
• Submit content you have no right to share, or another party's confidential information.
• Automate, scrape, overload, or otherwise abuse a feature, or attempt to extract or reverse-engineer the underlying models or prompts.

3. Accuracy and no warranty

AI-generated output may be incomplete, out of date, or wrong. It is provided “as is,” without warranty of any kind. You are responsible for verifying it before acting on it, and QAI is not liable for decisions you make based on AI output. This supplements the disclaimers and limitations of liability in the QAI Terms of Service.

4. Your data and AI providers

We process your inputs to provide the feature and as described in our Privacy Policy. Consistent with the commitments above, we minimize what we send to AI models, we do not train models on your data, and our providers process it transiently under enterprise terms that do not permit training on it. You retain ownership of your content and may request access, correction, or deletion at any time.

Feature-specific terms

The following clauses add to the general terms above for individual AI features.

5. QVC AI Feedback

This clause applies when you opt into instant AI Feedback on a QAI Ventures (QVC) pitch.

What it is: advisory, automatically generated commentary to help you sharpen your pitch — strengths, gaps, and questions to consider before you raise.

What it is not: an investment decision, funding offer, term sheet, due diligence, or professional advice. Receiving feedback does not mean QVC has reviewed, accepted, or declined your pitch, and no funding or admission decision is ever made solely by the AI.

What we send: the narrative parts of your pitch (problem, solution, market, traction, team, and, if attached, your deck text) are sent to a third-party large language model — currently Google's Gemini — to generate the feedback.

What we exclude: we send only a fixed, reviewed allowlist of fields and deliberately withhold your sensitive commercial data — exact raise amounts, valuation and cap-table percentages, monthly spend figures, your contact details, and any internal score. Those stay in your QVC record but are never part of the AI request.

Opt-in: AI Feedback runs only if you check its box at submission; checking it (together with accepting the Terms & Privacy Policy) is required to submit a pitch.

6. Changes and contact

We may update these AI Terms; material changes will be posted here with a new effective date. Questions? Contact us at support@qai.io.