US Tech Automations

Compliance paperwork AI disclosure notice pack 18 rows held

Companion chatbot notice — the rules behind it, quoted

The 6 published clauses that drive companion chatbot notices, from California, EU, New York, each with the publisher's own words and the case it does not reach.

Price
$49
Built for
A company running an AI chatbot, AI-generated content or an AI decision step that people in the EU or a US disclosure state can reach
Read
Re-read about every three months
Newest sealed read
2026-09-08

Email us for the $49 checkout linkNo card needed to ask. We reply with what we hold.

Newest sealed read: 2026-09-08. We re-read every source about every three months. We hold 1 sealed run going back to 2025-11-05.

What this page is18 rows held · newest sealed read 2026-09-08

  • 6 clauses across 3 places (California, EU, New York) drive this one notice, and they do not ask for the same thing.
  • Each clause below states who its own words reach and who they do not reach.
  • The draft text names the clauses it was written from, so you can read them rather than trust the wording.
  • Quoted passages behind these clauses: 12. Sources read 2026-09-08.

Real rows out of our sealed copies

These clauses all point at the same notice, from different places, and they do not ask for identical things.

6 rule clauses behind this notice sources read 2026-09-08
Rule and clauseWhat it asks forWho its own words reachWho they do not reachFrom when
EU AI Act, Regulation (EU) 2024/1689, Article 50(1)
Article 50(1)
Read it at the source
Tell the person they are dealing with an AI systemProviders of AI systems that interact directly with people — the Commission names chatbots, AI agents and avatars — wherever they are established, if the system is placed on the EU market, put into service in the EU, or its output is used in the EU.It does not reach a system that runs only in the background, only machine to machine, or with no direct contact with people; and it does not reach a case where it is obvious to a reasonably well-informed person that they are dealing with an AI.2 August 2026
California Business and Professions Code § 22602(a) (SB 243)
§ 22602(a)
Read it at the source
Say the companion chatbot is artificially generated and not humanOperators of companion chatbots, where a reasonable person interacting with it would be misled into believing they are interacting with a human.It does not reach a chatbot no reasonable person would take for a human, and the chapter is about companion chatbots rather than ordinary customer-support bots.1 January 2026
California Business and Professions Code § 22602(c) (SB 243)
§ 22602(c)(1)–(2)
Read it at the source
For a user the operator knows is a minor: say it is AI, and repeat every three hoursOperators of companion chatbots, for a user the operator knows is a minor: disclose the interaction is with artificial intelligence, and by default give a break reminder at least every three hours.It is written for users the operator KNOWS are minors, so it does not reach an adult user; the three-hour cadence is a default, not an absolute.1 January 2026
California Business and Professions Code § 22602(b) (SB 243)
§ 22602(b)
Read it at the source
Hold a self-harm crisis protocol and publish itOperators of companion chatbots: the bot does not engage with users at all unless the operator maintains a crisis protocol, and the details of that protocol are published on the operator's website.This is a protocol and a published page, not a notice line — a disclosure sentence does not answer it.1 January 2026
New York General Business Law § 1702 (Article 47, AI Companion Models)
§ 1702
Read it at the source
Say the user is not communicating with a human, at the start and every three hoursOperators of an AI companion: a clear and conspicuous notification at the beginning of the interaction, and at least every three hours in a continuing one, verbally or in writing.The start-of-interaction notification need not be given more than once per day, and Article 47 is written about AI companions rather than every chatbot.see the date note on this row
CITE-CHECK: the effective date of New York General Business Law Article 47
New York General Business Law § 1701 (Article 47, AI Companion Models)
§ 1701
Read it at the source
Hold a self-harm protocol that refers users to crisis servicesOperators of an AI companion: the companion contains a protocol for detecting and addressing suicidal ideation or self-harm, including a notification referring the user to crisis service providers.This is a protocol inside the product, not a line of notice text; a disclosure sentence does not answer it.see the date note on the § 1702 row
The draft text, watermarked drafted from 3 clauses
PartText
WatermarkDRAFT — buy to remove
The notice[your organisation] reminder: you are not communicating with a human. This companion is artificially generated and not human. If you are struggling, contact a crisis service — in the United States call or text 9-8-8.
WatermarkDRAFT — buy to remove
Where it goesAt the beginning of every companion interaction, and repeated in a continuing one at least every three hours.
Written fromny-1702, ca-comp-22602a, ca-comp-22602c

What this page cannot tell you

  • A notice is not the only thing some of these clauses ask for. Several ask for a protocol, a tool or a record, and no wording answers those.
  • The draft is a draft. This page does not say it is enough for any particular company, because that depends on facts it does not have.
  • We re-read the sources about every 90 days; past 180 days treat the page as stale.
  • We could not verify: the effective date of New York General Business Law Article 47.
  • Not affiliated with the European Commission, the California Legislative Counsel, the Colorado General Assembly, the Utah Legislature, the Maine Office of the Revisor of Statutes or the New York State Senate. Not legal, tax or professional advice. Rule text quoted from those publishers' own pages as of 2026-09-08.

See the file before you pay

You do not have to take our word for what is in the file. Here are 25 rows of the real thing, carrying all 9 of its columns, cut out of the dated copies we sealed ourselves. Nothing in it is made up and nothing in it is tidied up.

  • Open the 25 rows as a CSVA plain spreadsheet file. It saves to your machine rather than painting itself into a browser tab, and it opens in Excel, Numbers or Google Sheets.
  • The same 25 rows as JSONThe same rows again, laid out for reading with code.

What arrives after you pay: a private web page with every notice as plain text and as an HTML snippet, the dated matrix, the placement checklist and a downloadable bundle — within 15 minutes of payment. If it has not arrived, email operations@ustechautomations.com and a person sends it.

What you would be paying for: $49 once, for the AI disclosure notice pack. Nothing recurring, nothing to cancel and no subscription behind it. You get one private web page carrying each notice as copy-ready plain text and as a plain HTML snippet you paste into your own site, the dated matrix of every rule clause we hold with the publisher's exact words and a link to the page it came from, a checklist of where each notice goes and when it appears, the list of dates those rules name, and a JSON and text bundle you download and keep. You host the files; we host nothing for you. There is no badge, no public list of who bought it, no monitoring service and no promise to tell you when a rule changes. The rule text is quoted from the publishers' own pages and is dated: it is what those pages said on the day of the build, and the sources may lag. Four of the fifty-six quoted passages could not be re-checked from our machine and the pack names them. This is not legal, tax or professional advice, and nothing in it says whether your organisation complies with anything. Refund on request within 14 days. There is no pay button on this page yet. The private page is delivered within 15 minutes of payment. Still not here after 15 minutes? Reply to your Stripe receipt.

These 25 rows are a slice of the file, not the whole of it. What we cannot show you here is how far back it goes: that is the part you are paying for.

Buy the notice pack — $49

No pay button on this one yet. Email operations@ustechautomations.com. Ask what is in it before you pay. We reply with the current rule count, the list of what we could not verify, and the checkout link.

What you would be paying for: $49 once, for the AI disclosure notice pack. Nothing recurring, nothing to cancel and no subscription behind it. You get one private web page carrying each notice as copy-ready plain text and as a plain HTML snippet you paste into your own site, the dated matrix of every rule clause we hold with the publisher's exact words and a link to the page it came from, a checklist of where each notice goes and when it appears, the list of dates those rules name, and a JSON and text bundle you download and keep. You host the files; we host nothing for you. There is no badge, no public list of who bought it, no monitoring service and no promise to tell you when a rule changes. The rule text is quoted from the publishers' own pages and is dated: it is what those pages said on the day of the build, and the sources may lag. Four of the fifty-six quoted passages could not be re-checked from our machine and the pack names them. This is not legal, tax or professional advice, and nothing in it says whether your organisation complies with anything. Refund on request within 14 days. There is no pay button on this page yet. The private page is delivered within 15 minutes of payment. Still not here after 15 minutes? Reply to your Stripe receipt.

Email us for the $49 checkout link

No pay button on this page yet. The rule text is the publishers' own; the notice drafts are ours and are not legal advice.

More from this feed