Legal

AI Disclaimer

Always verify against the original FDA record.

Argus HQ — AI Disclaimer

Last updated: June 28, 2026 Version: 1.0

This disclaimer is incorporated by reference into the Terms of Service and the Privacy Policy. It is also displayed inside the product on the signup form, in every alert email footer, and in the dashboard footer.

Modeled on the AI usage disclaimers published by Notion AI, OpenAI ChatGPT Enterprise, Perplexity, and the FTC's 2024 guidance on AI-generated content disclosure.

Canonical disclaimer (use this exact text)

Argus HQ alerts are informational only. Summaries are generated in part by AI (Anthropic Claude) and may contain errors, misclassifications, or omissions. The underlying source documents (FDA, SEC, CPSC, etc.) are publicly available — always verify against the original source before making business, legal, or regulatory decisions. Argus HQ is not a law firm, accounting firm, or regulatory agency, and provides no legal, accounting, financial, medical, or regulatory advice.

This canonical text is the controlling version. Placement variants below restate it in shorter forms but the canonical text takes precedence for any interpretation question.

Placement variants

A. Signup checkbox (1 line)

Used as the label of the required consent checkbox on the signup form.

I agree to the Terms of Service, Privacy Policy, and AI Disclaimer. Argus HQ is informational only and not legal, financial, or medical advice.

B. Alert email footer (2 lines)

Appended to every alert email after the source citation, on its own block above the unsubscribe link.

Informational only. Summaries are AI-generated and may contain errors. Verify against the source URL above before acting. Not legal, financial, or medical advice. Unsubscribe · AI Disclaimer

C. Landing page footer (3 lines)

Displayed in the site footer beneath the link row.

Argus HQ is informational only. Alert summaries are generated in part by AI (Anthropic Claude) and may contain errors, misclassifications, or omissions. The underlying source documents (FDA, SEC, CPSC, Federal Register, SAM.gov) are public records. Always verify against the original source before acting. Argus HQ is not a law firm, accounting firm, or regulatory agency, and provides no legal, accounting, financial, medical, or regulatory advice.

D. Dashboard footer (1 line + link)

Persistent across all logged-in pages.

Argus HQ is informational only — not legal, financial, or medical advice. Full AI Disclaimer

E. Watchlist creation form (inline notice)

Displayed above the "Create watchlist" submit button on /watchlists/new.

Heads up: Argus HQ matches and summarizes public regulatory records using AI. We do not guarantee that every relevant filing will be caught, delivered on time, or summarized correctly. Verify every alert against the source URL before acting.

F. Cancellation flow (confirmation page, 1 line)

By cancelling, you confirm that any business decisions you made based on Argus HQ alerts were independently verified against the original source documents.

Why this matters (for internal / counsel reference, not customer-facing)

  1. FTC AI guidance (2024 onwards): the FTC has repeatedly warned that AI-generated content presented as authoritative without disclosure can be deceptive under §5 of the FTC Act. The canonical disclaimer is designed to satisfy that standard.
  2. State consumer-protection statutes: California, New York, and Texas have all increased scrutiny of AI-generated content in commercial settings. The verification-against-source language helps establish that customers are on notice.
  3. Negligent misrepresentation defense: repeated, prominent disclaimers help defeat reasonable-reliance arguments if a customer ever claims they suffered loss because Argus HQ "told them" something incorrect.
  4. Professional-services unauthorized-practice statutes: every US state has unauthorized-practice-of-law statutes. Explicit "not a law firm / not advice" language is the standard guardrail used by Casetext, Harvey, and legal-tech peers.
  5. EU AI Act (transparency obligations, Art. 50): AI-generated content interacting with natural persons must be disclosed as such. The canonical disclaimer meets that transparency obligation.

Internal rule (engineering)

  • The canonical disclaimer must be the single source of truth for AI-disclaimer text.
  • All placement variants live in lib/legal/ai-disclaimer.ts as named string constants (AI_DISCLAIMER_CANONICAL, AI_DISCLAIMER_SIGNUP, AI_DISCLAIMER_EMAIL_FOOTER, AI_DISCLAIMER_LANDING_FOOTER, AI_DISCLAIMER_DASHBOARD_FOOTER, AI_DISCLAIMER_WATCHLIST_INLINE, AI_DISCLAIMER_CANCEL).
  • Never compose disclaimer copy ad hoc in JSX. Always import from lib/legal/ai-disclaimer.ts.
  • When you change the canonical text, bump AI_DISCLAIMER_VERSION (used by the consent-logging table — see LEGAL_INTEGRATION_SPEC.md).