> For the complete documentation index, see [llms.txt](https://docs.moderncollections.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.moderncollections.io/compliance-for-integrators/ai-disclosure.md).

# AI disclosure

> **Not legal advice.** Get your own counsel.

## The trend

Several states have passed or are considering laws requiring **disclosure that an AI** (or "automated system") is interacting with a consumer. This is distinct from recording consent — it's about whether the human on the other end knows they're talking to a non-human.

What the platform's state-rules matrix (v1.2.0) actually flags as requiring AI disclosure today:

| State | Deadline                             | Status                                                                                           |
| ----- | ------------------------------------ | ------------------------------------------------------------------------------------------------ |
| CA    | 10 seconds (AB 2905, $500/violation) | Active                                                                                           |
| TX    | 30 seconds (SB 140)                  | Active                                                                                           |
| CO    | (no fixed deadline)                  | `ai_disclosure_stayed: true` — federal court stay April 2026, engine treats as P2 monitor-weekly |

Other state proposals (NY, NJ, CT, IL Video Interview Act, Utah SB 149) are tracked outside the engine and have not yet been encoded in the matrix. Federal: no comprehensive AI disclosure law yet, though FTC enforcement under "deceptive practices" has reached AI-non-disclosure cases.

## What our agent does

The voice agent always discloses at the start of every call that it is an AI assistant. The disclosure is rendered into the opening prompt for every flow — the canonical text is:

> "Hello, this is {{agent\_name}}, an AI assistant calling on behalf of {{creditor\_name}}."

When the compliance engine requires AI disclosure for the debtor's state, the prompt additionally renders a state-specific disclosure line that must fire within the state's deadline.

The agent's system card (v1.1.1) hard-codes that the agent never claims to be human and never denies being an AI; the same card forbids invented scarcity, specific social proof, and borrowed authority (FDCPA §807(1)/(3)/(9)). Credit-reporting language is rendered only for accounts where credit-bureau mention is enabled, and a post-call audit scan scores disclosure adherence on every call.

## What integrators need to do

### Email correspondence

We disclose AI involvement in email signatures and footers (in jurisdictions that require it). If you forward, reply to, or republish our emails on your side, **don't strip the disclosure**.

### Chat / portal flows

If you build a custom debtor-facing chat that's AI-driven, your side must do its own disclosure. We can't disclose for you.

### Misrepresentation risk

If you call the debtor on your own side and present yourself or our agent dishonestly ("I'm calling from \[Creditor's] collections team" without disclosing automation), you create exposure for both of us.

### Recordkeeping

In California and Texas, you may need to retain proof of AI disclosure in case of complaint. The audit trail records the rendered prompt version on every call (post-call extraction also scores adherence to the disclosure pause discipline). Pull the full audit timeline via `GET /v1/placements/{id}/audit`; the call record stores the `script_version` that was active at call time so transcripts can be replayed against the exact disclosure text.

## What to watch for

* New state laws — track via [NAAG](https://www.naag.org/), [NCSL](https://www.ncsl.org/), or a compliance subscription service.
* Federal proposals — likely accelerate in 2026-2027.
* Industry guidance — collections trade associations (ACA International, RMA) publish guidance.

## Anti-pattern alerts

| Anti-pattern                                                     | Why it's a problem                                                        |
| ---------------------------------------------------------------- | ------------------------------------------------------------------------- |
| Stripping the "AI assistant" line from forwarded transcripts     | Removes disclosure record; possibly creates a false impression downstream |
| Repurposing our voice in your own non-AI-disclosed channels      | Disclosure was for that one call; you can't reuse the consent             |
| Synthesizing the debtor's voice / behavior for training purposes | Increasingly restricted by state AI-training laws                         |

## See also

* [Recording consent](/compliance-for-integrators/recording-consent.md)
* [FDCPA for integrators](/compliance-for-integrators/fdcpa-for-integrators.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.moderncollections.io/compliance-for-integrators/ai-disclosure.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.