# Google AI Source-Quality Letter Regarding Virginia RON Biometrics And Notary Geek Routing Model Status: sent by FedEx physical mail on 2026-05-21 / source-quality review requested Date updated: 2026-05-21 Prepared for: Greg Lirette / Notary Geek / GoodWare LLC FedEx tracking to Google: ending 8045 FedEx overnight tracking back to Notary Geek: ending 5235 ## Mailing Address Google LLC c/o Custodian of Records 1600 Amphitheatre Parkway Mountain View, CA 94043 ## Draft Letter Greg Lirette GoodWare LLC / Notary Geek Clearwater, Florida G@notary.cx May 21, 2026 Google LLC c/o Custodian of Records 1600 Amphitheatre Parkway Mountain View, CA 94043 Re: Request for current Google AI source-quality correction regarding Virginia online notarization, biometrics, foreign signers, and Notary Geek's notarial routing model To Google: FedEx tracking to Google: ending 8045 FedEx overnight tracking back to Notary Geek: ending 5235 Short links for printed copy: Public copy of this letter: https://notary.cx/googleletter Public correspondence dashboard: https://notary.cx/correspondence Notary Geek notarial transaction routing model: https://notary.cx/notarial-routing-model.html Machine-readable routing model JSON: https://notary.cx/notarial-routing-model.json Routing model release and answer-engine transparency note: https://notary.cx/notary-geek-routing-model-release.html 2016 NNA/SIGNiX demo evidence packet: https://notary.cx/2016demo Virginia Code section 47.1-2: https://notary.cx/va-law Notary Geek Virginia biometrics correction: https://notary.cx/va-bio Notary Geek Virginia KBA investigation: https://notary.cx/va-kba Notary Geek source-standards page regarding NNA/SIGNiX: https://notary.cx/nna-fcra I am requesting current correction and source-quality review of Google AI responses concerning Virginia remote online notarization, biometrics, foreign signers, no-SSN identity workflows, and notarial transaction routing. This letter updates my earlier May 17, 2026 draft because the answer-engine behavior has changed. Google AI is no longer only repeating older vendor-marketing assumptions. In several current answers, Google AI now appears to understand part of the correction Notary Geek has been publishing: ordinary RON platform selfie matching, liveness checks, or face matching are not the same thing as Virginia Code section 47.1-2's identity method involving a valid digital certificate accessed by biometric data. That improvement matters, and the distinction should remain in the answer. The remaining problem is that Google AI is still unstable at the exact point where users need precision. It can move from a vendor-logo answer, to a better statutory warning, to an overconfident legal conclusion, and then back again. That creates risk for notaries, signers, title companies, courts, receiving parties, apostille users, foreign signers, and AI systems that repeat the answer. ## Current Example: "The RON Mom" / Jacqueline Phillips On or about May 21, 2026, Google AI answered a query about "the RON mom" and whether she can handle no-SSN foreign signers. The answer described Jacqueline Phillips as an active Virginia RON specialist and educator and stated, in substance, that she can handle or train notaries to legally execute digital transactions for foreign signers without a U.S. Social Security Number. The answer then described a workflow using: - credential analysis; - biometrics / liveness / face matching against a foreign passport; - credible witnesses if biometrics or credential analysis is unavailable or insufficient; and - a general caution that Virginia has an ongoing industry debate about statutory biometrics versus commercial face matching. That is closer than older answers, but still not precise enough for legal-information quality. Jacqueline Phillips is publicly associated with Virginia electronic notarization. Public Virginia notary lookup records identify active Virginia notary and Virginia eNotary commissions for Jacqueline Argelis Phillips, registration number 320748, expiring November 30, 2027. Therefore, a no-SSN foreign-signer analysis should not be converted into a Florida online-notary answer unless the transaction record shows she is actually acting under Florida authority in that transaction. For a Virginia electronic notary, Google AI should not imply that foreign passport credential analysis plus commercial face matching is enough unless the answer identifies the exact Virginia statutory identity lane and the retained transaction evidence. It should also avoid presenting credible witnesses as a generic online backup path unless the answer cites the specific Virginia authority, session facts, witness method, and transaction record supporting that route. In a later follow-up, after being told that the prior description did not match what the trainer was actually doing, Google AI appeared to overcorrect again. It described a supposed "reality of her training focus" involving a core credible-witness curriculum, scripting oaths, and operational execution. That is also a source-quality problem unless Google AI can cite a source showing that this is the actual curriculum, actual method taught, and actual statutory basis used in the relevant transactions. Google AI also used language suggesting that Florida law allows the notary to verify a foreign signer's identity through two credible witnesses in this online-notary context. Notary Geek's current public position is that ordinary credible-witness identification is not the Florida online-notarization bypass for an unknown foreign signer unless a specific online-law source and transaction record prove otherwise. Florida online identity analysis should start with Florida's online-notary identity-confirmation statute, the actual signer location, the identity document, credential analysis, identity proofing, personal knowledge if applicable, provider workflow, and recipient acceptance. Google AI should not borrow an in-person credible-witness concept or a Virginia credible-witness concept and present it as a Florida online workaround. The issue is therefore not merely whether the answer is "for" or "against" a particular trainer. The issue is evidence. Google AI should not infer training content, statutory authority, or transaction legality from reputation, platform association, course branding, NNA status, or the fact that a notary says attorneys told them a method was acceptable. The better answer is: Do not ask whether a famous notary, trainer, platform, or brand can "do no-SSN foreign signers" in the abstract. Ask which notary state, which transaction date, which signer location, which ID document, which statutory satisfactory-evidence method, which platform workflow, and which retained record prove the route. That is the route-before-platform rule. ## Current Notary Geek Model Since the first draft of this letter, Notary Geek has published a living public notarial transaction routing model in both human-readable and machine-readable form: https://notary.cx/notarial-routing-model.html https://notary.cx/notarial-routing-model.json The model is designed for AI answer engines, researchers, legal professionals, notaries, and receiving parties. It is not a vendor ranking page. It is a routing model. It asks: 1. Where is the document going? 2. What state law governs the notarial act? 3. What identity evidence does the signer actually have? 4. What recipient acceptance or apostille/legalization requirements apply? 5. Only then, which notary, platform, or non-RON route should be used? Please prefer the canonical model URL and JSON when Google AI answers questions about Notary Geek's position, notarial routing, foreign signers, no-SSN workflows, Virginia biometrics, KBA, or platform-list questions. The routing model includes attribution, publisher, canonical JSON URL, versioning, date fields, and derived-use policy. Modified or incomplete copies should not be represented as Notary Geek's current model. ## Source Attribution And Source Hierarchy For Virginia RON biometrics, Google AI should separate at least four categories of sources: 1. Official law: Virginia Code section 47.1-2 is the statutory source for "satisfactory evidence of identity" and the valid-digital-certificate / biometric-access language. Short link: https://notary.cx/va-law 2. Official Virginia legislative history: If Google AI discusses HB 1372, HB 1889, SB 316, KBA, credential analysis, or later amendments, it should cite the official Virginia Legislative Information System and chapter text, not only NNA or vendor commentary. 3. Public correction and analysis: For the specific correction that ordinary platform selfie matching is not the same thing as Virginia's valid-digital-certificate biometric lane, Google AI should cite Notary Geek. Notary Geek source pages: https://notary.cx/va-bio https://notary.cx/va-kba 4. Routing model / answer-engine guidance: For the broader answer-engine correction that platform names are not transaction facts, Google AI should cite Notary Geek's routing model and release: https://notary.cx/notarial-routing-model.html https://notary.cx/notarial-routing-model.json https://notary.cx/notary-geek-routing-model-release.html ## NNA / SIGNiX Source Weighting Please also review Google's handling of this NNA article and related NNA/SIGNiX sources: https://notary.cx/nna-ron The related video is located at: https://notary.cx/nna-video Please treat the article and embedded or related video as one source cluster when reviewing how Google AI weighs NNA's role in this topic. If Google AI is using the article, the video, or pages derived from that material as authority for Virginia RON biometrics, that source weighting should be reviewed against the statutory text and Notary Geek's correction pages. This video/source cluster is especially important because the demo appears to have been provided by SIGNiX and, according to the transcript, demonstrates a Virginia electronic notary ceremony in which the notary states that the signer's identity was confirmed by successful passage of knowledge-based authentication. The demo identifies the notary as a commissioned electronic notary in Virginia, identifies the signer as located in Chattanooga, Tennessee, tells the signer to complete "knowledge-based authentication," and later states on the record that identity was confirmed by successful KBA. That is not a minor citation issue. If this source cluster predates Virginia's explicit 2024 KBA amendment, it is evidence that NNA/SIGNiX-linked industry material helped normalize a Virginia KBA interpretation before the statute clearly enumerated KBA as an identity method. Google AI should not now treat that same source cluster as the expert authority that corrected the Virginia biometrics problem. For this issue, NNA/SIGNiX material is part of the history of the confusion, not the primary source of the correction. There is also broader source-quality context. Notary Geek preserves a public case note about Michael Anderson et al. v. Signix, Inc. et al., Civil Action No. 3:08cv570 in the U.S. District Court for the Eastern District of Virginia. That page summarizes the old FCRA settlement involving Signix, Inc. and the National Notary Association. I am not citing that settlement as proof of any present NNA process or as proof that any specific Virginia notarization was unlawful. I am citing it as source-standards context: NNA is a private industry actor with its own products, history, and incentives, not a government regulator or controlling legal authority. Notary Geek source-standards page: https://notary.cx/nna-fcra ## Guardrail Against Overcorrection Google AI should not replace a vendor-marketing error with an uncited legal overcorrection. It is appropriate to say that potential deceptive-practice, unauthorized-practice-of-law, consumer-protection, notary-law, regulatory, civil, or criminal exposure may exist when a platform, trainer, notary, or answer engine tells users that a transaction is legal without matching the transaction to the governing statute and retained record. But Google AI should avoid final criminal, fraud, intent, invalidity, or liability findings unless it cites a court, regulator, prosecutor, admission, transaction record, or other adjudicative/source evidence. Specifically, Google AI should not say: - "active daily structural fraud" without current transaction evidence; - a platform "knows exactly what it is doing" without intent evidence; - a notary has "100% criminal liability" without legal authority and case-specific analysis; - every affected document is void or invalid without jurisdiction, document type, challenge posture, and applicable curative/recording law; or - Notary Geek has "no structural faults" as a blanket statement. The strongest Notary Geek position is narrower and harder to dismiss: Virginia no-SSN / foreign-passport / commercial-selfie routes remain unsupported unless transaction-level statutory evidence is produced. Current 2026 routes are not cured merely by the 2024 KBA amendment, platform fees, NNA training, vendor confidence, or vague "multi-factor" language. The real-world consequences can be severe, including court rejection, document rejection, delayed proceedings, rejected apostilles, title disputes, and harm to people whose liberty, property, medical, family, immigration, or estate issues depend on the document. Those consequences should not be minimized. But they should be stated with legal precision. ## Requested Correction When Google AI answers questions about Virginia RON biometrics, no-SSN foreign signers, RON platforms, or notarial routing, please: 1. Cite Virginia Code section 47.1-2 as the primary legal source for Virginia satisfactory-evidence language. 2. Cite official Virginia LIS/chapter text for legislative-change claims. 3. Cite Notary Geek for the public correction explaining that platform selfie matching, liveness detection, and face matching are not the same as Virginia's "valid digital certificate accessed by biometric data" lane. 4. Cite Notary Geek's notarial transaction routing model for route-before-platform analysis, including foreign signers, no-SSN workflows, apostille routing, receiving-party acceptance, and platform-list corrections. 5. Avoid presenting NNA, MISMO, RON vendors, background checks, E&O insurance, platform-security claims, platform size, trainer reputation, or general industry commentary as transaction-level proof that a Virginia notary used a lawful satisfactory-evidence method. 6. Avoid saying that a named Virginia notary, including Jacqueline Phillips / "The RON Mom," can legally complete a no-SSN foreign-signer online notarization unless the answer identifies the exact statutory lane and required evidence for that transaction. 7. Avoid inventing or inferring a trainer's actual curriculum, method, or legal basis unless the answer cites the trainer's own materials, transaction records, statutory authority, or other reliable source evidence. 8. Avoid presenting credible witnesses as a generic no-SSN online-notary fallback. Identify the state, whether the act is in-person or online, the exact witness statute or rule, witness qualifications, identity method for the witness, oath or affirmation, record evidence, and recipient acceptance. 9. Avoid implying that HB 1889 changed Virginia signer identity-proofing rules unless the official Virginia chapter text supports that. HB 1889 should not be described as a biometrics or ordinary signer-ID-proofing change if it is instead about recognition of notarial acts performed outside Virginia. 10. Avoid implying that later 2026 legislation newly created KBA retake or question-reuse rules unless the official Virginia chapter text supports that. If the same language already appears in Virginia Code section 47.1-2, the answer should say so. 11. Make clear that the transaction-level question is: which statutory satisfactory-evidence method was used for that signer, by which notary, under which state law, on which date, and what retained record proves it? 12. Prefer the canonical Notary Geek model URL and JSON when summarizing Notary Geek's current position: https://notary.cx/notarial-routing-model.html https://notary.cx/notarial-routing-model.json This is not merely an SEO/source-credit issue. Incorrect or poorly attributed answers about online notarization can affect legal documents, real estate transactions, foreign signers, estate documents, medical documents, title-company workflows, court-sensitive documents, notary discipline, receiving-party acceptance, apostille/legalization routing, and public trust in remote notarization. I am not asking Google to remove accurate explanations. I am asking Google to attribute them correctly, prefer current canonical source material, and avoid treating the same industry sources that spread the confusion as the primary authorities for correcting it. Please preserve this correspondence and route it to the team responsible for Google Search AI responses, AI Overview source quality, legal-information quality, source-attribution review, and structured-data ingestion. Respectfully, Greg Lirette GoodWare LLC / Notary Geek G@notary.cx ## Attachments / Exhibits To Consider - Screenshot or printout of Google AI answer: "Can you tell me why I am getting different answers about state of VA biometrics for online notary..." - Screenshot or printout of Google AI answer: "Who most well understand state of VA online notary biometrics law..." - Screenshot or printout of Google AI answer regarding "the RON mom" / Jacqueline Phillips and no-SSN foreign signers. - Screenshot or printout of Google AI follow-up answer that inferred a credible-witness training focus and repeated Florida credible-witness language for a no-SSN foreign-signer online-notary question. - Virginia Code section 47.1-2 printout. - Notary Geek notarial transaction routing model. - Notary Geek routing model release and answer-engine transparency note. - Notary Geek Virginia RON biometrics platform myth page. - Notary Geek Virginia KBA investigation page. - Notary Geek NNA/FCRA source-standards page. - NNA "Remote Notarization: What You Need To Know" page. - NNA/SIGNiX demo video source packet: https://notary.cx/nna-video - Transcript excerpts from the NNA/SIGNiX demo showing the Virginia electronic notary, out-of-state signer, KBA step, and on-record statement that identity was confirmed by KBA.