Fusion Collective Sentinel Citation

Fusion Sentinel Citation

Citation integrity.

Leverage statistical analysis with multiple LLMs to verify and score every citation in your federal filings, before you submit. Then run the same analysis on opposing counsel's.

Your first filing is free. No card required.

Coverage

Federal District, Circuit, and Supreme Court filings

Sources

CourtListener, Cornell LII, eCFR, Scholar

Foundation

Patent-pending (US 64/056,937)

Why this exists

Mata v. Avianca, 2023: six fabricated citations, sanctions on the docket. In re Prince Global Holdings, 2026: around twenty-eight erroneous citations on an emergency motion by one of the country's most prestigious firms, same district, three years later. The lesson nobody needed to learn twice has been learned several times.

The tooling has to be different now.

Run Sentinel Citation on the brief you're drafting. Then run it on opposing counsel's. The report tells you what's fabricated and what's weak. Your first filing is on us.

01 / The problem

Citation review cannot be a vibe check.

You can't trust just one. A single model's intuition is not evidence. A naive citation checker passes a citation through a single language model and asks "does this look real?" That answers the wrong question. The right question is whether the cited authority exists, whether you read it accurately, and whether your treatment of it matches what the court actually said.

02 / The method

Four phases. No shortcuts.

Phase 01 / Extract

Find every citation

Three language models read the filing independently and return every legal citation they identify. We take the union of their output and combine it with Eyecite — never relying on any single model's word for it.

Phase 02 / Verify

Confirm each one exists

Each unique citation is checked against authoritative sources: CourtListener for case law, Cornell LII for statutes and rules, eCFR for regulations, with a Scholar fallback. Anything that doesn't resolve is flagged, not silently dropped.

Phase 03 / Judge

Read the authority in context

With the cited authority's text fetched, three judging models — Claude Sonnet, Gemini Flash, and GPT, each from a rival vendor — evaluate the filing's use of it across the four checks. The multi-model jury is the core of the patent-pending methodology.

Phase 04 / Synthesize

One verdict per citation

A fourth model, the meta-judge, reconciles the three verdicts into a final score and reasoning for each citation, then writes a filing-level summary — something a partner can read in two minutes.

Claude Sonnet GPT Gemini Flash Meta-judge

03 / The numbers worth knowing

Why hand checks miss this.

Purpose-built legal AI tools still err on 17%+ of queries — because a single model checking citations is still a single model. Sentinel was built on the opposite premise: no single model's word is ever taken for anything.

58%

Hallucination rate, generic LLMs on legal queries.

Stanford "Large Legal Fictions" · Dahl et al. 2024

17%+

Error rate in purpose-built legal AI tools.

Stanford RegLab "Hallucination-Free?" · Magesh et al. 2024

~28

Erroneous citations on a single emergency motion.

In re Prince Global Holdings · S.D.N.Y. Bankr. 2026

04 / The difference

Why this is different.

Single-LLM citation check Sentinel Citation
Approach One model, one prompt, one opinion. Three models extract independently. Verdicts reconciled by a fourth meta-judge.
Verification "Does this look like a real citation?" Each citation looked up in CourtListener, Cornell LII, or eCFR. Authority text fetched.
Treatment Inferred from the citation alone. The cited authority is read and compared against the filing's use of it.
Output A pass or a flag. Per-citation scores across the four checks, plus a filing-level narrative.
Failure mode Silent miss when the model is overconfident. Misses surfaced as "not found" or "not judged" with a reason.
Audit trail The model said so. Every verdict carries the source it was checked against and the model's reasoning.

05 / Four checks per citation

What we actually verify.

Three of these protect you from hallucinations and bad lawyering. The fourth gives you something to attack with. Run the analysis on your own filings before you submit — and on opposing counsel's after they do.

01 · Existence

Every cited case and statute is cross-referenced against authoritative sources — CourtListener for case law, Cornell LII for statutes and federal rules, eCFR for regulations. If the cited authority does not exist, the citation is flagged before anything else happens.

02 · Quote Accuracy

Anything the filing presents as a quotation is matched against the cited authority's actual text. Misquotations, paraphrases dressed up in quote marks, and outright fabrications surface here.

03 · Holding Support

The cited authority is read in full and compared against the proposition the filing offers it for. We name the gap when the case does not actually say what the brief says it says.

04 · Citation Strength

Above and beyond hallucination defense. Each surviving citation is rated on how well it supports its proposition: binding versus persuasive, current good law versus narrowed or criticized, on-point versus distinguishable. On your filings it surfaces soft spots before a clerk does; on opposing counsel's, before oral argument.

06 / The strategic read

Even a clean brief earns its fee.

Suppose the report finds zero fabrications. You still walk away with a Citation Strength map of the entire filing: which citations carry the argument, which are binding versus merely persuasive, which would fold under a motion to strike — in your brief, and in theirs. The fabrication check is why you run it the first time. The strategic read is why you run it every time.

See the deeper evidence — coverage, consistency, and the Westlaw question →

A position, not a promise

Citation review is the litigator's job.

We are not interested in a product that hides behind "AI did it." Citation review is your name on the filing. Sentinel Citation does the parts that don't reward judgment: extracting every citation, verifying each against authoritative sources, reading the cited authority alongside your use of it. The judgment stays with you. Every output is traceable, every verdict has a reason, every model that touched the filing is named in the report.

The question

Would you sign a brief whose citations you didn't verify?

The cases exist or they don't.

They say what you think they say or they don't.

The court will check. So should you.

Your name still goes on the filing.

See a verified filing

See exactly what Sentinel sees.

These are complete, unedited Sentinel Citation reports run on real public court filings — including filings that later drew sanctions. Where a judge ruled on the filing, we show you the court's words next to Sentinel's verdicts. No demo data. No cherry-picking. This is the report you get.

Pricing

Not another subscription.

$99 a filing — or fund a balance and pay only for what you analyze. Flat or prepaid; your first filing's free, no card required.

Frequently asked

Questions we hear from litigators.

No. Sentinel Citation is the first pass. It catches fabrications, misquotes, and treatment errors that humans miss when fatigued. The final read is yours, every time.
Yes. The same analysis runs on any filing you can read. On your own briefs you get the report before you submit; on opposing counsel's you get a map of where the citations are weakest. Citation Strength scoring is included on every analysis.
Initial release covers US federal court filings: District, Circuit, and Supreme Court briefs, motions, and memoranda. State court coverage is on the roadmap.
A fixed three-plus-one ensemble: three judging models from rival vendors — Claude Sonnet (Anthropic), Gemini Flash (Google), and GPT (OpenAI) — plus a meta-judge that synthesizes their verdicts. The exact versions are pinned and recorded on every run, so every verdict is attributable. The composition isn't user-configurable: cross-vendor model diversity is what gives the verdicts statistical weight. The report names every model that contributed to each verdict, so you can see the spread.
CourtListener for case law, Cornell LII for statutes and federal rules, eCFR for regulations, and Google Scholar as a fallback for opinions available only behind Westlaw or Lexis paywalls. We do not invent authority.
A typical 30-page brief runs in a few minutes. Longer filings scale linearly because verification calls and model calls run in parallel.
It lands in a "not judged" bucket with the reason ("CourtListener returned not-found," "no opinion text available," and so on). It is not silently dropped, and it is not silently passed. See how not-found citations are triaged →
Your filing is encrypted in transit and at rest, read by software rather than staff, and never used to train models. The analysis is retained in your account, and you can delete a filing and its analysis at any time. Read the security & confidentiality page →
A flat $99 per filing — no subscription, no seats, no contract. Your first filing is free and needs no card. See pricing →
Sentinel is our broader evaluation engine for AI-generated work, built on patent-pending drift-detection methodology (US 64/056,937). Sentinel Citation is the same engine, specialized for legal filings. If you already use Sentinel, Citation runs in the same workspace.

When you are ready

See what a verified filing looks like.

Pricing is straightforward and starts at the per-filing level. There's no demo gate, and your first filing is on us.

Foundation

Patent-pending (US 64/056,937)

Coverage

Federal District, Circuit, Supreme Court filings

Methodology

Three judging models + one meta-judge

Sources

CourtListener, Cornell LII, eCFR, Scholar