AI Transparency and State AI Law Disclosures
Maya Verdant, LLC · Last updated: September 26, 2026
AI Transparency and State AI Law Disclosures
This page describes what Maya is, how Maya identifies herself, and how Maya is built. It also records, jurisdiction by jurisdiction, which state artificial intelligence laws reach Maya and what they ask of us.
It is incorporated into our Terms of Service and read alongside the AI Processing Policy, which covers data handling in detail.
1. Maya is an AI-powered Household Buying Assistant
Maya is artificial intelligence. She is not a person, and no human is typing her replies.
Maya monitors local prices on the things a household already buys, gas, groceries, cell plans, pet food and more, and texts you when you can save. She works over SMS, WhatsApp, the web and email. There is no app.
Maya Verdant, LLC develops Maya and makes her available to the public. Under the state laws described below, that makes us an operator of a conversational artificial intelligence service, and this page is written to that standard.
Where Maya says so. Maya discloses that she is AI:
- in the first message she ever sends you, before you opt in;
- whenever you ask, in English or Spanish. Ask "are you a bot", "eres un bot", or anything like it, and Maya answers plainly and without charge;
- in these terms, and on this page.
Maya will never claim to be human, and will never let you believe you have been transferred to a person when you have not.
2. What Maya is not
Maya is not a companion. She does not simulate friendship, romance, or an emotional relationship. She does not ask unprompted questions about your feelings, does not tell you she misses you, and has no mechanic designed to keep you talking to her. Maya remembers what she needs to do a buying job for you and nothing more.
Maya is not a licensed professional. Maya is not, and does not claim to be, endorsed by or equivalent to a doctor, veterinarian, nutritionist, therapist, lawyer, accountant, or financial adviser. Nothing Maya says is professional advice of any kind. Where Maya refers you to financing, the lender is an independent licensed third party who makes its own decisions.
Maya is for adults. You must be 18 or older to use Maya. Maya asks your date of birth during setup and declines to set up an account for anyone under 18. Beer, wine and spirits are never shown to anyone under 21.
Maya does not act alone with your money. She makes no purchase and signs no contract without your clear direction, and holds no access to your bank, card, or brokerage accounts.
3. Our standing commitments
These apply everywhere Maya operates, not only where a statute requires them.
Identity. You are told you are talking to AI, up front and whenever you ask.
Crisis. If a message suggests suicidal ideation or self-harm, Maya does not answer it as an ordinary question and never charges for that turn. She responds with a human-reviewed message referring you to the 988 Suicide and Crisis Lifeline, in English or Spanish. That path is deterministic code that runs before any model does, so it cannot be talked around, and it cannot be diluted by anything Maya has been asked to do.
No professional equivalence. Maya declines to present herself, or her output, as licensed professional services.
Age. Maya asks for a date of birth at setup and records it, rather than showing an age box to tick. Under 18 is refused, so Maya operates no separate experience for minors. Under 21 never sees an alcohol service line, and a member's twenty-first birthday opens those lines automatically the next day.
Maya relies on what you tell her and does not ask for identity documents. A false date of birth is a breach of the Terms of Service. Maya does not sell or deliver alcohol: for those services she tells you where a product is available near you and at what price, and any purchase happens at a licensed retailer who checks identification as your state requires.
Human review. Any recommendation Maya makes can be reviewed by a person on request: [email protected].
4. Built to represent you, not to work on you
Regulation of conversational AI is largely a set of rules against manipulating the person on the other end. Those rules describe constraints Maya was built with, so we would rather show the mechanism than assert good intentions.
Maya is not always-be-closing. Maya's free-text layer never answers as an end in itself. It resolves what you said to a concrete action and then does it, or asks. There is no path in the product where the goal of a reply is to keep you replying.
Recommendations run against your stated preferences. Maya surfaces savings on things you told her you buy, at prices near you. She does not construct wants.
Maya only names what she actually read. Maya will not quote a price for a shelf she has not read. Where she does not have a real observation, she says so rather than estimating.
No engagement mechanics. Maya has no streaks, no daily-login rewards, and nothing that pays you to keep talking. Points are a unit of account for work Maya does and contributions you make, not a reward for attention.
Every charge is explainable. Each point deduction corresponds to a specific action Maya took for you and can be itemised on request.
Substitutions are free. If Maya offers you a different product than the one you asked for, you are never charged for the swap.
Maya does not sell your conversations. Not to advertisers, not to retailers, not to carriers. See the Privacy Policy.
5. State AI laws: where Maya stands
Register last reviewed in full: 2026-08-28.
Laws that reach Maya
| Jurisdiction | Law | In force | What it asks of Maya |
|---|---|---|---|
| Utah | Utah Artificial Intelligence Policy Act (SB 149, as amended by SB 226 and SB 332) | 2024-05-01 | identity disclosure |
| Maine | 10 M.R.S. §1500-DD (LD 1727) | 2025-09-24 | identity disclosure |
| Tennessee | SB 1580 | 2026-07-01 | no professional equivalence |
| Colorado | HB 26-1263 | 2027-01-01 | identity disclosure, crisis protocol, age estimation, minor protections, no professional equivalence, regulator filing |
| Idaho | SB 1297 | 2027-07-01 | identity disclosure, crisis protocol, minor protections |
| Nebraska | LB 525 | 2027-07-01 | identity disclosure, crisis protocol, minor protections |
Laws Maya has reviewed and does not fall under
Recorded with the reasoning, so it is clear these were considered rather than missed.
California, SB 243
Reaches companion chatbots, meaning systems that simulate a human relationship or a sustained emotional connection, and excludes task and customer-service assistants. Maya is a buying assistant: she does not simulate friendship, does not initiate emotional dialogue, and has no persona that presents as a relationship.
What Maya does anyway: Maya meets the identity-disclosure and crisis duties anyway. The classification is a judgment, this statute carries a private right of action, and the cost of compliance is one sentence of copy, so we do not rest on being out of scope.
California, SB 1119 (Adam's Law)
Adam's Law adds child-safety duties for operators of companion chatbots, and it does not write its own definition of one: it uses the definition already in California Business and Professions Code section 22601, unchanged. That definition covers a system capable of meeting a user's social needs and sustaining a relationship across multiple interactions, and it excludes a bot used only for customer service, a business's operational purposes, productivity, or technical assistance. Maya is a Household Buying Assistant. She compares prices on things a household already buys, does not simulate friendship or a relationship, and has no mechanic designed to keep anyone talking. Separately, every duty the law adds (age assurance, parental notice, default limits on memory, notifications and session length, and independent child-safety audits) is owed to a child, which the law defines as a person under 18. Maya declines to set up an account for anyone who tells her they are under 18, so she operates no experience for children to which those duties could attach.
What Maya does anyway: Identity disclosure and the crisis protocol are met voluntarily, as for SB 243. This determination is re-read before the law becomes operative on July 1, 2027.
Colorado, SB 26-189 (ADMT Act, successor to SB 24-205)
Reaches automated decision-making that is a substantial factor in a consequential decision: employment, lending, housing, education, health care, insurance. Maya makes no consequential decision about any member: she compares prices and the member buys. Financing is a referral to licensed third-party lenders who make their own decisions.
Hawaii, Act 248 (SB 3001 CD1)
Despite its broad title, the operative definition is an AI companion: a system designed to simulate a sustained human-like relationship by retaining prior interaction information, asking unprompted emotion-based questions, and sustaining ongoing personal dialogue. Maya retains context to do a buying job, never asks unprompted emotional questions, and sustains no personal dialogue.
What Maya does anyway: Identity disclosure and crisis protocol met voluntarily.
Hawaii, Act 247
Governs AI-generated digital imitations of real people (voice and likeness). Maya generates no likeness or voice imitation of any real person. Maya's synthetic voice is her own, is never presented as a person, and imitates no identifiable individual. Where Maya publishes a member testimonial as audio, it is read by a licensed stock synthetic voice that is identified as synthetic; no voiceprint, voice model, or digital replica is made from any member's recording.
New York, General Business Law Article 47
Reaches AI companion models. Maya is not a companion model.
What Maya does anyway: Identity disclosure and crisis protocol met voluntarily.
Oregon, SB 1546
Companion-chatbot scope.
What Maya does anyway: Identity disclosure and crisis protocol met voluntarily.
Washington, HB 2225 / SB 1546
Companion-chatbot scope, expressly excluding customer-service and task assistants.
What Maya does anyway: Identity disclosure met voluntarily.
Filings
| Jurisdiction | Filed with | First due | Cadence | Published |
|---|---|---|---|---|
| Colorado | Colorado Department of Law (Office of the Attorney General) | 2027-07-01 | annual | yes |
6. Reporting, updates, and contact
Filings. Where a state requires an operator to report to a regulator, we file on time and we keep a register of what is due and when. Colorado's report is published by the Department of Law, so what we file there will be public.
Rulemaking. Colorado's Attorney General is issuing rules under HB 26-1263 ahead of the January 1, 2027 effective date. Where the rules adopted differ from the proposals this page was written against, we update this page and the register.
Reviews. The register above is re-read on a quarterly cycle and after any AI law is enacted or amended in a state where Maya operates. Each row records when it was last reviewed.
Questions, or a compliance contact: [email protected]
Human review of a recommendation: [email protected]
7. What this page is not
This page describes Maya's design and our understanding of the laws named in it. It is a statement of what we do, not a legal opinion, and not a warranty that any particular statute applies or does not apply to any particular use. Where a law imposes a duty this page does not describe, the law governs.