AI VORTEXLEGAL AI RISK
AI visibilityConsultingUse free
LIVECorpus refreshed Aug 9, 2026 · Latest decision Aug 6, 20261,868 records · 283 US courts · 53 US jurisdictionsMethodology
Global evidence map1,868 matched records · 40 mapped countries · geographic clusters synchronize the case list
Global legal AI risk matters by countrySelect a country cluster to inspect its evidence snapshot. Page scrolling remains available over this embedded map. Cluster size represents tracked record count and color represents the highest editorial impact classification.1Tanzania, 1 tracked records, highest editorial impact classification low206Canada, 206 tracked records, highest editorial impact classification career-ending1297United States, 1297 tracked records, highest editorial impact classification career-ending1Papua New Guinea, 1 tracked records, highest editorial impact classification medium8Argentina, 8 tracked records, highest editorial impact classification career-ending3Chile, 3 tracked records, highest editorial impact classification career-ending1Kenya, 1 tracked records, highest editorial impact classification low1Russian Federation, 1 tracked records, highest editorial impact classification high1The Bahamas, 1 tracked records, highest editorial impact classification medium6South Africa, 6 tracked records, highest editorial impact classification high10Brazil, 10 tracked records, highest editorial impact classification high3Colombia, 3 tracked records, highest editorial impact classification high1Costa Rica, 1 tracked records, highest editorial impact classification low13France, 13 tracked records, highest editorial impact classification medium1Zimbabwe, 1 tracked records, highest editorial impact classification low55Israel, 55 tracked records, highest editorial impact classification high2United Arab Emirates, 2 tracked records, highest editorial impact classification high1Qatar, 1 tracked records, highest editorial impact classification low4South Korea, 4 tracked records, highest editorial impact classification low13India, 13 tracked records, highest editorial impact classification medium2Poland, 2 tracked records, highest editorial impact classification low2Austria, 2 tracked records, highest editorial impact classification low8Germany, 8 tracked records, highest editorial impact classification low8Belgium, 8 tracked records, highest editorial impact classification high7Netherlands, 7 tracked records, highest editorial impact classification high1Portugal, 1 tracked records, highest editorial impact classification high8Spain, 8 tracked records, highest editorial impact classification high6Ireland, 6 tracked records, highest editorial impact classification low7New Zealand, 7 tracked records, highest editorial impact classification low97Australia, 97 tracked records, highest editorial impact classification career-ending1China, 1 tracked records, highest editorial impact classification low15Italy, 15 tracked records, highest editorial impact classification high1Denmark, 1 tracked records, highest editorial impact classification high61United Kingdom, 61 tracked records, highest editorial impact classification high2Czech Republic, 2 tracked records, highest editorial impact classification high1Northern Cyprus (KKTC), 1 tracked records, highest editorial impact classification low1Uganda, 1 tracked records, highest editorial impact classification low1Trinidad & Tobago, 1 tracked records, highest editorial impact classification high
Map colorCareer impactHighMediumLowCluster size = tracked record count

Prepare your next move
with evidence you can trace.

Review your work, challenge a weak claim, compare judicial responses, or investigate a pattern across source-linked legal AI records.

Try a search:
Select a country to inspect its evidence snapshotCountry directory

One evidence layer. Three source-backed actions.

Search the record

Discover cases, courts, tools, consequences, and failure patterns across the public corpus.

Start searching
Inspect the source

Open the recorded document, understand its source tier, and see what the available evidence cannot establish.

Review methodology
Build the review

Turn a case or filtered analytics view into a source-complete, print-ready evidence brief.

Open analytics
Start from the work

Built around the decision you need to make.

The same public record should behave differently for a filing team, chambers, a researcher, or a product operator. Choose the workflow closest to yours.

Litigation team

Check a filing before it leaves the firm

Run a citation, quotation, proposition-support, and disclosure review with a reusable verification record.

Open the filing gate Court or chambers

Inspect the record without inferring AI use

Separate demonstrated discrepancies from attribution, preserve source links, and compare similar public matters.

Open neutral review Research and knowledge

Trace a pattern across courts and sources

Search the global corpus, inspect the denominator, and open the underlying public evidence before relying.

Explore risk topics Legal tech and risk

Measure the record without overstating it

Inspect denominators, missing values, evidence coverage, and the records behind every displayed pattern.

Open public analytics

Recent & significant records

View all records
Withers v. City of AberdeenN.D. Mississippi · Jun 8, 2026fake citations / fabricated quotesCareer impact$8.0KLnu v. Blanche9th Cir. CA · Jun 3, 2026fake citations / fabricated quotesCareer impact$5.0KCouvrette v. WisnovskyD. Oregon · Mar 23, 2026fake citations / fabricated quotesCareer impact$110.2KJoel A. Rivera v. Triad Properties Corporation, et al.N.D. Alabama · Mar 31, 2026fake citations / fabricated quotesCareer impact$35.6KIn the Matter of Janelle Melissa Lewis, an attorney and counselor-at-lawApp. Div. 1st Dept (NY) · May 14, 2026fake citations / immigrationCareer impact$1.5KScott v. Illinois Human Rights Comm’nCA Illinois · Jul 28, 2026fake citations / fabricated quotesHigh$15.0K
Corpus intelligence

See the record, its shape, and its limits.

Counts describe tracked public matters, not industry incidence. Failure and consequence tags can overlap. Monetary metrics use only records with a known amount.

Open full analytics
Tracked matters1,86842 countries · Apr 14, 2023 to Aug 6, 2026Source-linked99.8%1,865 of 1,868 records contain a source URLCourts represented62453 US jurisdictions representedLatest 30 days112Relative to latest recorded decision
Corpus growthCumulative record and monthly additionsInteractive decision-date view
View

Decision month. Corpus-added dates are not recorded. The latest period may be incomplete. Hover for values; click a period to inspect records.

Observed consequencesWhat the record says happened
tags overlap
Warning or admonishment626
Monetary sanction392
Bar or professional referral149
Professional discipline136
Dismissal or adverse disposition111
Recurring signalsLeading failure modes
tags overlap
Fabricated authorities1,590
Trial1,574
Pro Se1,096
Misrepresented authority793
Fabricated or altered quotations616
ConcentrationMost represented courts
corpus coverage
S.D. New York48
SC New York35
E.D. Michigan30
W.D. Washington30
C.D. California29
1,854 non-allegation-only records 216 known monetary amounts3 missing source linksRead methodology
Trust boundary

Every answer should show where the evidence ends.

Source-linked, not source-equivalent. A link establishes traceability; it does not make a summary a substitute for the order or docket.

Observed incidents, not failure rates. Counts are not adjusted for product usage, filing volume, or reporting bias.

Neutral attribution. A discrepancy is not proof that AI was used. AI Vortex states the recorded basis and flags what remains unknown.

Inspect methodology, coverage, and limitations
PUBLIC INTELLIGENCE STAYS FREE

The record should remain open, inspectable, and useful.

Search the complete public corpus, inspect recorded sources, open analytics, and generate AI Vortex-attributed evidence briefs without an account.

Public research

Free / no account required

  • Search every public matter and recorded source
  • Explore courts, jurisdictions, topics, and analytics
  • Generate source-complete briefs with AI Vortex attribution
  • Submit corrections and inspect evidence limitations
Search the public record
HELP SHAPE THE PUBLIC LAYER
Founding review group

Review / test real records

  • Test search quality and record clarity
  • Flag missing sources or classification issues
  • Review print briefs and evidence boundaries
  • Influence what the public product supports next
Join the review group

No endorsement implied. Reviewers help improve accuracy and usability.