Skip to main content
Citation Safe

Free tool · no signup

AI Hallucination Risk Self-Test

Ten honest questions about your actual AI-drafting workflow. Scoring is plain arithmetic — every point value is shown below, nothing is hidden in a model. This is not a statistical prediction, not a percentile ranking against other firms (we don’t have that data), and not legal advice.

1. Which AI tool(s) do you use to draft or research filings?
2. Are AI-drafted citations checked against a primary source before filing?
3. If citations are checked, who does it?
4. Roughly how many filings per month include AI-drafted or AI-assisted text?
5. Which practice area do you draft AI-assisted filings in?
6. Do you cite unpublished opinions, out-of-jurisdiction cases, or less common sources with AI assistance?
7. Does your AI drafting tool cite sources with a clickable link/reference you can verify, or plain text only?
8. Has a citation in one of your filings ever been flagged or challenged by a court or opposing counsel (for any reason, AI or not)?
9. If AI drafts a citation you don't personally recognize, what's your default action?
10. Does your jurisdiction have a standing order or local rule on AI disclosure/citation certification, and do you follow it?

Answer all 10 questions to see your score.

Also try

  • Tax Cite Safe — IRC sections, Treasury Regulations, and Tax Court citations
  • Med Cite Safe — Medical literature citations — PubMed, Crossref, ClinicalTrials.gov, with retraction flagging
  • FAR Check — FAR, DFARS, GAO bid-protest, and Federal Circuit/COFC citations

Need more than one? See the Argus Verification Suite (3-Pack $69/mo, All-Access $99/mo, Firm $299/mo).

This self-test is a directional exercise built from public patterns in reported AI-citation sanctions cases, not a statistical model trained on outcome data, and not legal advice. Example cases link to our sanctions database.