# FreeWIS third-party AI disclosure and permission

**Effective for FreeWIS native v1 in the United States · Adopted September 16, 2026.**

## Short disclosure

**The FreeWIS AI assistant uses Anthropic's API to generate a response.** If you continue, FreeWIS
will send Anthropic your message and the relevant conversation context needed to
answer it. Relevant context may include excerpts retrieved from the FreeWIS
knowledge base and, inside PLAY, results from tools operating on synthetic
demonstration records.

**Do not send real case or personal information.** Do not include a real
person's name with case facts, contact details, identifiers, health or financial
information, abuse or neglect reports, or real-data files, screenshots, images
or audio. Use general questions or fictional examples only.

The AI assistant is artificial intelligence, not a human or professional adviser. Output
may be inaccurate, incomplete or unsuitable. It does not search the open web.
Standalone CHAT uses HCT-curated knowledge-base materials; PLAY may also use its
offered synthetic demonstration tools. Check current official sources when
currency matters. Do not use AI output for real-case decisions, professional advice,
emergencies or child-abuse reports.

The [FreeWIS Privacy Notice](https://freewis.com/privacy) identifies HCT's
purposes, providers, storage and retention qualifications and links Anthropic's
published practices. The [Terms of Service](https://freewis.com/terms) and the
applicable [PLAY Agreement](https://freewis.com/agreement) or [CHAT access
notice](https://freewis.com/chat-access) also apply.

## Expanded information

### Information sent for a response

Depending on the request and feature used, FreeWIS may send:

- the person's message or attached synthetic content;
- prior messages or instructions included as relevant conversation context;
- retrieved FreeWIS knowledge-base excerpts;
- inside PLAY, results from approved tools operating on synthetic demonstration
  data; and
- technical request information needed to operate, secure and troubleshoot the
  response.

### What permission means

This action permits the described third-party processing needed to generate AI-assistant responses. It
does not authorize real-case input, advertising, sale, general-purpose model
training on identifiable submissions, an unrelated new use, a payment, PLAY
admission, standalone CHAT admission, source-code access or an Agency action.

Permission remains subject to the current Privacy Notice and may be presented
again when a material provider, purpose or data-flow change requires it.

### Provider and retention boundary

Anthropic's actual commercial account, enabled features and applicable terms
determine its processing and retention. HCT does not promise zero provider
retention or simultaneous deletion of every provider-held copy. FreeWIS's own
content-retention rules, protected backups, incident evidence and legal holds
are described in the Privacy Notice.

### Reporting an output

Use the in-app **Report AI output** action to report or flag an AI response.
Before submission, the action identifies the conversation excerpt and technical
context that will be attached. It is not an emergency or child-abuse reporting
channel.

FreeWIS records the account/provider-scoped identity, disclosure version and
hash, choice and timestamp, client/service context and related legal-document
versions needed to honor and document the choice. A permission record is
separate from whether a later message is sent to Anthropic or a response
succeeds.

**Hudson Consulting and Training, LLC** · Mailing address: 7765 Wadsworth Blvd #740433, Arvada, CO 80006 ·
**freewis.support@hudsonct.com**