AI & Automation Disclosure

This AI & Automation Disclosure explains how Vantis uses artificial intelligence to generate certain outputs, what data is and is not sent to the AI provider, and the limitations of AI-generated content. It supplements but does not modify the Terms of Service (especially §§6, 11, 12, 13, 18, 19, 20) and the Privacy Policy (especially §§3, 3.3, 3.4). In any conflict between this document and the Terms of Service, the Terms of Service control. The Disclaimers page is a non-binding plain-English summary; in any conflict between the Disclaimers page and any other Vantis legal document, the other document controls.

0. Status of Vantis

Vantis LLC is a Washington-based software company that publishes an analytics and research tool for self-directed investors. Vantis is not a registered investment adviser under the Investment Advisers Act of 1940, the Washington Securities Act (RCW 21.20), or any state investment-adviser law; is not a broker-dealer; is not a robo-adviser; is not a financial planner; does not custody assets; does not access your brokerage account; and does not place trades on your behalf. The use of artificial intelligence to generate narrative commentary in the Service does not change any of the above.

1. What Vantis Generates with AI

Some outputs you see in the Service include narrative commentary generated by a third-party large language model. Currently the provider is Anthropic (Claude). Specifically:

• Per-ticker analysis thesis — a short qualitative note that summarizes the quantitative signals for a given ticker (e.g., why implied volatility is elevated, what the put-skew suggests, etc.). This text is generated on demand or as part of a scan.
• Covered-call commentary — when you run a covered-call analysis, an optional narrative paragraph describes the structural mechanics of strikes that score highest under our methodology applied to public data. The paragraph is generic to the ticker; it does not see your share count, cost basis, account size, tax position, target price, or any uploaded position data, and it is not a proposal that you sell any specific call against any specific position.

AI-generated text appears alongside its source quantitative data and is identified as AI-generated. We do not present AI-generated commentary as though it were not AI-generated.

When AI commentary is unavailable — the model is not configured, temporarily rate-limited, over a cost limit, or returning an error — the Service does not silently substitute or fabricate AI output. It shows the underlying scores and data without AI commentary, and may show a summary generated directly from the same published quantitative signals and gates.

The quantitative scores, signals, and gates are computed deterministically and do not depend on the AI model; only the narrative commentary uses AI.

2. What Data Is Sent to the AI

Each prompt sent to Anthropic includes only:

• the ticker symbol you are analyzing;
• public market data for that ticker (price, implied volatility, historical volatility, options Greeks, fundamental ratios, recent options flow);
• current market context (e.g., VIX level, scheduled Federal Reserve meeting dates, scheduled earnings dates); and
• Vantis’s generic methodology context.

The prompt does not include your name, email address, IP address, account identifiers, subscription state, payment information, session data, portfolio holdings, account balances, trade history, journal entries, watchlists, position quantities, cost basis, target prices, or any other Content you have uploaded. The prompt is constructed to use only the ticker, public market data, and our generic methodology context. Anonymization is verified by an audit when the AI model or prompt templates materially change, and on a periodic basis. We retain audit records for up to two years and, to the extent they remain available, provide them to regulators on request. We do not train any AI model — internal or external — on Content you upload.

Anthropic processes the prompt as a service provider to Vantis and is contractually prohibited from training its models on Vantis API traffic. Anthropic’s published practices indicate that prompts and responses are deleted within 30 days, except where Anthropic is required by law to retain them longer or its automated trust-and-safety classifiers flag the prompt — in which case Anthropic may retain and review the prompt and response for an extended period. Anthropic does not human-review prompts in the ordinary course; the trust-and- safety carve-out is the exception. See Privacy Policy §3.3 and trust.anthropic.com.

2.1 Disabling AI Thesis Generation

AI thesis generation is off by default for new accounts; you affirmatively enable it at first use. The always-anonymized prompt is described in §2 above. You can disable AI processing at any time at Settings → Privacy.

When AI thesis is disabled, the Service still computes scores, structure candidates, and gates locally — only the narrative paragraph generated by Anthropic is omitted. You can re-enable at any time. Each toggle is recorded in our internal audit log (timestamp, IP, user-agent) so we can verify the setting was honored.

Toggle records in our audit log are retained for 24 months for security and dispute-resolution purposes. They are not transmitted to Anthropic or any third-party AI provider, and are not used for marketing, profiling, or any automated decision affecting you. You may request deletion as described in the Privacy Policy.

Disabling AI thesis stops new prompts from being sent. Prompts already sent to Anthropic before you disabled the feature are subject to Anthropic’s standard retention (see Privacy Policy §3.3) and cannot be retroactively recalled.

3. Limitations of AI-Generated Content

You should know:

• Factual errors and hallucinations. AI commentary may contain factual errors, omissions, or hallucinations, including numerical errors, fabricated relationships between data points, incorrect attributions of cause, and invented references. You must not act on any AI-generated factual claim — including any number, ratio, percentile, directional statement, or relationship between data points — without independently verifying it against an authoritative source. Vantis does not warrant the accuracy, completeness, timeliness, or fitness for any purpose of any AI output. (See Terms §§11, 18.)
• Confident prose ≠ accuracy. AI text may read like the work of a professional analyst even when wrong. Tone and fluency are not evidence of accuracy.
• No human review before display. Vantis does not human-review individual AI outputs before you see them. We periodically sample outputs for quality but do not pre-vet them.
• Anthropic-side review. Anthropic’s automated trust-and-safety classifiers may flag prompts for extended retention and human review at Anthropic. See §2 above and Privacy Policy §3.3.
• Stale knowledge. The AI’s training data has a cutoff date. The AI may not know about recent events, corporate actions, regulatory changes, or other developments.
• Model changes. We may upgrade, downgrade, replace, or remove the AI provider at any time. Different models produce different outputs. We will update this Disclosure if we change providers, and we re-audit the prompt after each material model change to confirm the data-flow commitments in §2 still hold.
• Illustrative references. AI commentary may reference specific tickers, options structures, strike prices, expirations, or scenarios. Any such reference is illustrative of our methodology applied to public data and is not a recommendation, an endorsement, or an offer to engage in any specific transaction. Modeled or hypothetical performance referenced in any AI output is not indicative of future results, does not reflect actual trading, and may not account for fees, slippage, taxes, assignment risk, or other real-world costs.
• Not investment advice. AI commentary is informational text. Vantis is not a registered investment adviser, broker-dealer, or financial planner, and Vantis is not acting in any fiduciary, advisory, or planning capacity when AI commentary is generated or displayed.

4. What the AI Does Not Do

The AI does not:

• Recommend trades for you to take
• Place orders or interact with your brokerage
• Assess whether any trade is suitable for your situation, risk tolerance, or goals
• Have visibility into your portfolio or account
• Predict future market behavior or returns
• Replace human professional advice

Nothing displayed by the Service is offered, marketed, labeled, or held out by Vantis as a “recommendation” to you. Where the Service uses words like “highest-scoring,” “top candidate,” or similar, those phrases describe a methodology output applied to public data, not a recommendation that you take any action.

4.1 No Advisory or Continuing Relationship

The display of AI commentary does not create — and is not intended to create — any investment-adviser, broker-dealer, fiduciary, planning, agency, or similar relationship between you and Vantis. Vantis owes you no duty of care, no duty of loyalty, no suitability obligation, and no duty to monitor, update, or correct any AI commentary after it is generated or displayed. AI commentary is one-shot, ticker-generic, and produced without knowledge of you, your account, your goals, your tax position, your risk capacity, your time horizon, or any of your other holdings.

4.2 Apparent Personalization

Where the Service displays AI-generated commentary in proximity to data you have uploaded, the AI is rendering a description of the data you supplied, not an analysis of your individual financial circumstances. Apparent personalization is a function of the data you choose to upload; it is not advice tailored to you, and no advisory or fiduciary relationship arises from the proximity of any AI output to your data.

5. No Reliance

AI commentary is provided for informational and educational purposes only. It is not investment advice, a recommendation, an endorsement, an offer, a solicitation, or a suitability assessment, and is not personalized to you.

You must not use AI commentary, in whole or in part, as a basis for any investment decision. You must independently verify every factual statement, number, ratio, percentile, and directional claim against an authoritative source before acting on it. AI text often appears in confident, declarative prose; that confidence is a feature of how language models produce text and is not evidence of accuracy, expertise, or analytical rigor.

By using the Service you waive any claim that AI commentary induced you to take or refrain from taking any investment action (see Terms of Service §13). If you require investment advice, consult a licensed professional who has reviewed your full circumstances.

6. Changes & Disclosure of New Uses

If we begin using AI in new ways (for example, to assist customer support, to summarize portfolio data, or to generate other types of content), we will update this Disclosure before the new use goes live. Material changes will be communicated via the email address associated with your account.

A version history of legal documents (including this one) is published at /legal/changelog.

7. Contact

Questions about AI use at Vantis: privacy@getvantis.ai