AI Disclaimer
Last updated: 15 August 2026
AI output can be confidently wrong. Language models can misread a chart, invent a price level, or produce plausible-sounding analysis with no basis. Verify everything against your own chart and broker before acting on it.
VANTIQX uses artificial intelligence to analyse charts, answer questions, generate strategy code and summarise markets. This page explains what that means, where it can go wrong, and what happens to what you submit.
1. Which AI providers we use
Depending on the feature, prompts and any images you submit are processed by Google (Gemini API), Anthropic (Claude API), and for some features Groq and Hugging Face. Requests are routed through our own server so that our provider keys stay server-side. We do not use your content to train our own models.
2. What you send
Chart screenshots and other images you upload for analysis are transmitted to the AI provider handling that request, together with your prompt. Do not upload identity documents, account statements, private correspondence, or anything else you would not want processed by a third party.
3. Limits of the output
- AI-generated analysis is a probabilistic reading of the input, not a calculation and not a fact.
- Models can hallucinate levels, misidentify an instrument or timeframe, and misread a low-resolution or cropped screenshot.
- Model knowledge has a cutoff and no awareness of events after it; live prices come from market-data providers, not from the model.
- The same question can produce different answers on different runs.
4. Confidence scores and technical readings
Numeric confidence scores, sentiment gauges and indicator consensus readings in the app are computed from price data and from the indicators' own outputs. They describe how strongly the calculated inputs agree at that moment — they are not a probability that a trade will win, and they are not a measure of how other traders are positioned.
5. Generated strategy code
Strategy or indicator code produced by the AI Strategy Builder is a starting point, not a finished product. Review it, test it on historical data, and forward-test it on a demo account before risking real money. You are responsible for anything you run.
6. Not advice
Nothing produced by AI in VANTIQX is financial advice or a personal recommendation. See the Risk Disclaimer.
7. Human oversight
In Manual mode — the default — you remain the decision-maker. Every order requires your explicit confirmation, and an AI signal or analysis never places a trade by itself.
In Autopilot mode that changes, and it changes only because you change it. Once you give explicit, recorded consent and arm Autopilot on an account, VANTIQX's automated decision engine does place trades automatically toward your goal — inside hard limits enforced in code, not by the engine. Your oversight then works through the consent gate, the limits you set, a notification for every order, an audit log of every order and every refusal, and an instant halt. See the Order Execution & Automation Disclosure.
8. Contact
Questions about AI processing: contact@tadjfze.com.