How much plaintiff intake verification is enough? A tiered checklist for class action and mass arbitration sign-ups
Court records show most claims in some recent settlements were fraudulent or invalid. What Wallrich and arbitration rules expect for each claimant, and a tiered intake checklist.
Plaintiff intake verification works best in tiers: quiet, automatic checks on every sign-up, documented proof before you retain someone or put them forward as a class representative, and a file that ties each claimant to the facts before anything is filed in their name. Settlement records now show fraudulent claims arriving by the million. And in Wallrich v. Samsung Electronics America, Inc. (7th Cir. July 1, 2024), counsel-signed demands and a spreadsheet of names and addresses did not show that any of 35,651 consumers were Samsung customers who had agreed to arbitrate.
- In four of the five settlement records we reviewed, most submitted claims were flagged as fraudulent or found invalid; in JUUL, 77.45% of 14.5 million claims scored high for fraud indicia.
- In Wallrich v. Samsung, the Seventh Circuit held that counsel-signed demands and a spreadsheet of names did not show 35,651 consumers were Samsung customers.
- AAA rules call for the representative's affirmation that each case's information is true and correct to the best of their knowledge; JAMS calls for counsel's sworn declaration.
- Settlement administrators told the CFPB that claims rates fell nearly 90% when documentary proof was required.
Why does intake verification matter more than it used to?
An open web form lets a real person sign up in two minutes and a script sign up thousands of times. Settlement administrators now put numbers on that in sworn declarations.
In the JUUL economic loss settlements, administrator Epiq told the court it received 14,488,748 claims and scored 77.45% of them high for fraud indicia (declaration, April 24, 2024, ¶¶ 7 to 8). Judge William Orrick approved a plan for "rejecting claims that are clearly fraudulent without expending class resources unnecessarily" (order, May 7, 2024).
JUUL is not an outlier:
| Settlement and court | Submitted | Valid or allowed | Also in the record |
|---|---|---|---|
| JUUL economic loss settlements, N.D. Cal. | 14,488,748 claims (Apr. 2024) | 842,004 valid claimants (court order, Oct. 2024) | 77.45% of claims scored high for fraud indicia |
| Chicco KidFit booster seats, Jimenez (later Sayers) v. Artsana USA, S.D.N.Y. | 3,333,633 claim forms for 28,658,800 products as of Oct. 30, 2023 (claims closed Jan. 2024) | 65,174 products validated (plaintiffs' filing, Feb. 2025) | Plaintiffs reported about 8.6 million fraudulent or invalid claims; the judge noted 875,000 products could qualify |
| Nixon v. Grande Cosmetics, D.N.J. | 6,526,866 claim submissions (plaintiff's filing, Mar. 2024) | 120,232 valid claims (court order, July 2024) | Grande had estimated the class at about 1,000,000 consumers |
| Hesse v. Godiva Chocolatier, S.D.N.Y. | 829,637 claims (Mar. 2022) | 435,230 valid (Godiva's letter, Dec. 2022) | Godiva's letter said Kroll found a bot "stemming from a foreign country" |
| In re Novartis and Par Antitrust Litigation, S.D.N.Y. | 132,192 consumer claims | 22,342 allowed | 107,540 flagged as potentially fraudulent (Dec. 2023) |
Sources: JUUL declaration and distribution order; Artsana declaration, hearing transcript and final approval brief; Grande filing and final approval order; Godiva declaration and letter; Novartis declaration. Units differ by case (claims, claimants, products). Compiled by ClassAct, October 25, 2026.
These are settlement claims, not firm intake, but an intake form is an open web form too. The Novartis administrator said that in her experience, settlements with an open class, no documentation requirement and no cap on the amount claimed tend to draw a "high proportion of potentially fraudulent claim submissions" (declaration, December 20, 2023, n.3).
Mass arbitration raises the stakes per name, because each demand is its own case. The AAA's 2025 data release (June 3, 2026) gives "whether the claimants met any contractual pre-conditions to filing, which arbitration agreement applies, and whether the claimants met the relevant filing requirements" as typical issues for the Process Arbitrator.
What do arbitration rules and courts expect you to show about each claimant?
Less at filing than at the first real challenge.
The AAA's Mass Arbitration Supplementary Rules (amended and effective April 1, 2024) say "A separate Demand for Arbitration must be filed in each individual case." The filing party also submits an intake spreadsheet that goes to the other side, plus "an affirmation that the information provided for each individual case is true and correct to the best of the representative's knowledge" (MA-2). If the AAA appoints a Process Arbitrator, disputes over filing requirements and how to cure deficiencies can go there (MA-6(a), (c)(i)).
The JAMS Mass Arbitration Procedures (effective May 1, 2024), where the parties agreed to them, say "A Demand for Arbitration Form and applicable arbitration agreement must be submitted for each Claimant" (Proc. 2(a)). They also call for "a sworn declaration from counsel" to the same best-of-knowledge standard (Proc. 2(c)).
Wallrich shows what comes after intake. The consumers had asked a court to compel arbitration, so they carried the burden, and Samsung conceded its customers were bound. The question became whether each claimant was "in fact" a Samsung customer.
The court noted that "The AAA's filing requirements involve no substantive determinations" and that the demands were "nothing more than allegations." It said the consumers "could have submitted almost anything to meet their burden," such as "receipts, order numbers, or confirmation numbers," or declarations. Its summary-judgment-like posture "does not allow second chances" (slip op. 12 to 15).
Some limits: Wallrich binds federal courts in Illinois, Indiana and Wisconsin, it applied the AAA's earlier multiple-case filing rules rather than the 2024 version, and a separate holding addressed fees. Your clause, forum and circuit may differ.
How much proof can you ask for before real members drop out?
Some, but each added step can cost real people. In its 2015 arbitration study, the CFPB reported that five administrators of consumer financial settlements said procedural complexity affected claims rates: "Claims rates fell nearly 90% if documentary proof was required" (Section 8, p. 31).
The FTC's 2019 staff report was more cautious. It did not find documentation requirements related to claims rates overall (p. 3). A footnote reports that forms asking for the most difficult documentation, such as original receipts or notarizations, had claims rates "an average of 27.3 percentage points lower," a marginally significant result, and only 5% of cases asked for that kind of proof (p. 39; p. 45 n.61). The two studies measure differently, so their figures don't compare directly.
Courts look at the other side of the ledger too. In Novartis, the court referred an inquiry "to assess the reasons for the disproportionately low number of claims allowed compared to claims filed," then adopted a finding that the process "has integrity" (order, May 4, 2026). A written screening rule, applied the same way to everyone, is easier to explain than case-by-case calls.
The workable answer is to ask for proof when it changes a decision, and not before.
What should a plaintiff intake verification checklist cover at each stage?
A sign-up costs the firm little; a filed demand carries a fee and an affirmation.
| Stage | Check | What it catches | Cost to real sign-ups |
|---|---|---|---|
| 1. Sign-up | Yes or no knockouts: state of residence (a required field), date range, product or account | Ineligible and out-of-state sign-ups | Low |
| 1. Sign-up | Duplicate match on email, phone, and name plus address | Repeat sign-ups and serial filers | Low (households share details) |
| 1. Sign-up | Traffic signals: many submissions from one IP address, identical answers, very fast completions | Bots and scripted traffic | Low (shared networks can trip it) |
| 1. Sign-up | A plain note on what proof you'll ask for later, with an optional upload now | Surprise drop-off at stage 2 | Low |
| 2. Before engagement or naming a rep | Proof of purchase or account: receipt, order or confirmation number, statement line, screenshot | Non-customers | Medium to high |
| 2. Before engagement or naming a rep | Name, address and contact details match the proof; a live conversation with every rep candidate | Typos, impostors, unwilling reps | Low |
Sources: AAA rules, MA-2 (Apr. 1, 2024); JAMS procedures, Proc. 2(c) (May 1, 2024); Wallrich, slip op. 12 to 15; FTC data guidance ("Keep only what you need"). Compiled by ClassAct, October 25, 2026.
A worked example
Take a hypothetical California-only claim against a subscription app, with a week of ads behind it.
- Sign-up. Three knockouts: lived in California during the class period, paid for the subscription, paid within the date range. Duplicate and traffic checks send flagged entries to a review queue rather than rejecting them.
- Shortlist. The firm asks for one thing: a screenshot of the receipt email or billing page showing name, plan and date.
- Filing. If the terms carry an arbitration clause, each file gets the account email, sign-up date, where the person subscribed (app or web can mean different terms), any required notice of dispute and a signed declaration.
How do you keep real sign-ups engaged while you check them?
Verification adds waiting, and waiting loses people.
- Use a second channel. The court-approved JUUL plan called for verification notices by email and, where one bounced, a second attempt by alternate email or text message (order, May 7, 2024). For intake, the equivalent is a channel the person agreed to at sign-up.
- Repeat the ask. In the FTC's settlement sample, cases that tried to reach class members more than once had "average and median claims rates that are more than twice as high" as single-attempt cases (p. 27). Our claims rate breakdown covers notice methods.
- Ask for one thing at a time, with an example of acceptable proof.
What this means for your firm
- Write down each case's knockouts and fraud flags before the ads run, and keep the version you used.
- Ask for proof where it changes a decision: engagement, naming a representative, filing.
- For mass arbitration, compare each file with the evidence Wallrich found missing, and read it alongside the clause and current forum rules.
- Send borderline flags to review. In JUUL, the court approved rejecting clearly fraudulent claims outright and sending the middle tier to verification.
- Collect only what each stage needs.
Rules vary by jurisdiction. Check the rules that apply to you.
FAQ
Checking that a person who signs up for a class action or mass arbitration is real, is who they say they are, and fits the case's criteria, such as state, dates, product or account. One approach works in stages: automatic checks at sign-up, then documented proof before retaining someone, naming a representative or filing a demand.
The Seventh Circuit held in 2024 that a spreadsheet of only names and addresses failed to show that the claimants were Samsung customers, and that counsel-signed demands were allegations, not evidence. It said receipts, order or confirmation numbers, or declarations could have met the claimants' burden.
It can, though the research covers settlement claims, not litigation sign-ups. Five administrators told the CFPB in 2015 that claims rates fell nearly 90% when documentary proof was required. The FTC's 2019 study found no overall link and only a marginally significant drop for the 5% of cases requiring original receipts or notarizations.
The FTC's guidance for businesses says "Keep only what you need for your business" and advises against collecting sensitive identifying information without a legitimate need. At sign-up that often means contact details and knockout answers, with receipts and declarations requested later.
Sources
- Wallrich v. Samsung Electronics America, Inc., No. 23-2842, 106 F.4th 609 (7th Cir. July 1, 2024), slip opinion
- American Arbitration Association, Mass Arbitration Supplementary Rules, amended and effective April 1, 2024
- JAMS, Mass Arbitration Procedures and Guidelines, effective May 1, 2024
- American Arbitration Association, 2025 mass arbitration data release, June 3, 2026
- In re JUUL Labs, Inc., No. 3:19-md-02913 (N.D. Cal.), Epiq declaration on fraudulent claims, ECF No. 4218-1, April 24, 2024
- In re JUUL Labs, Inc., order approving fraud plan, ECF No. 4224, May 7, 2024
- In re JUUL Labs, Inc., filing reporting verification results, ECF No. 4325, October 4, 2024
- In re JUUL Labs, Inc., order authorizing distribution, ECF No. 4326, October 8, 2024
- Jimenez v. Artsana USA, Inc., Angeion declaration on claims received, ECF No. 97-1, November 6, 2023
- Jimenez v. Artsana USA, Inc., hearing transcript, ECF No. 105, November 8, 2023
- Jimenez v. Artsana USA, Inc., plaintiffs' final approval memorandum, ECF No. 133, February 7, 2025
- Nixon v. Grande Cosmetics, LLC, No. 1:22-cv-06639 (D.N.J.), filing on claim submissions, ECF No. 52, March 7, 2024
- Nixon v. Grande Cosmetics, LLC, final approval order, ECF No. 72, July 18, 2024
- Hesse v. Godiva Chocolatier, Inc., No. 1:19-cv-00972 (S.D.N.Y.), Kroll declaration, ECF No. 105, March 21, 2022
- Hesse v. Godiva Chocolatier, Inc., Godiva letter on fraudulent claims, ECF No. 155, December 16, 2022
- In re Novartis and Par Antitrust Litigation, No. 1:18-cv-04361 (S.D.N.Y.), Angeion declaration, ECF No. 648, December 20, 2023
- In re Novartis and Par Antitrust Litigation, order adopting report and recommendation, ECF No. 670, May 4, 2026
- Federal Trade Commission staff report, Consumers and Class Actions: A Retrospective and Analysis of Settlement Campaigns, September 2019
- Consumer Financial Protection Bureau, Arbitration Study: Report to Congress, March 2015
- Federal Trade Commission, Protecting Personal Information: A Guide for Business, October 2016
ClassAct is a technology platform, not a law firm. This post is general information, not legal advice.



