Optional AI Terms

Effective: September 9, 2026 Version: ai-2026-09-09-v2 Provider: OpenAI API

These terms are between Viva Longer LLC, doing business as Ashby Care (“Ashby”), and the childcare organization (“Center”) whose owner separately accepts them. They supplement the Center’s existing Terms of Service, Privacy Policy, and Data Processing Addendum (“DPA”) only for optional AI writing help. Accepting Ashby’s general terms does not enable AI or accept these AI terms.

1. Your choice

AI is off by default, including for existing Centers. The owner may enable it during onboarding or later in Settings → AI writing help by separately accepting the current AI terms. The rest of Ashby remains available if the Center declines. Enabling AI applies to authorized staff across the Center’s organization, including its locations; it does not automatically generate or send anything.

Ashby records the owner’s account identifier, the agreement version, the choice, and its timestamp. A new version of these AI terms requires fresh owner acceptance before further AI drafting. This is the Center’s instruction for this optional service, not a substitute for any parental consent or notice otherwise required for the Center’s activities.

2. What we send

Only when an authorized staff member requests a draft, Ashby sends OpenAI selected activity categories (painting, reading, music, outdoor play, blocks, crafts, dance, puzzles, gardening, pretend play, science, or storytelling), a generic message purpose, or a school-announcement topic (closure, event, supplies, meeting, payment, or policy), together with a tone choice and drafting instructions. Activity drafts may also include a short or detailed length choice. OpenAI returns proposed text.

Ashby does not send child names, dates of birth, profiles, medical or incident details, photos, image URLs, or message conversations. Unsupported free text is blocked. Where needed, a permitted child reference is restored inside Ashby after generation. This agreement does not authorize sending sensitive child information to an AI provider.

Announcement dates, times, event choices, supply selections, and owner-approved school facts stay inside Ashby and are inserted after generation. School facts are structured opening days and hours, pickup instructions, and supplies approved by the owner for a location. Ashby stores those facts in its own database; it does not send them to OpenAI. Staff must check that the selected facts are current and appropriate for the recipients.

Reusable saved wording is a separate, local feature that works with AI off. Each item is private to its author within the Center and is not sent to OpenAI. Ashby treats both its title and body as sensitive content. Items expire 90 days after saving and are then unavailable for reuse; expired items are removed by Ashby’s retention cleanup, and also when the author’s list is next read or saved. The author may delete an item sooner. Deleting or expiring saved wording does not delete a copy already used in a message or report, which follows that record’s retention rules. “Local” here means stored inside Ashby, not only on the author’s device.

3. OpenAI processing and retention

The Center authorizes OpenAI as a subprocessor solely for this optional drafting service. OpenAI processes the permitted inputs and generated drafts under its API terms and applicable data-processing terms. Ashby does not opt this content into model training or voluntary data sharing.

OpenAI does not use API inputs or outputs to train models by default. Its abuse-monitoring logs may contain inputs and outputs and are generally retained for up to 30 days, unless longer retention is required by law or reasonably necessary to protect its services or third parties. This is standard retention, not Zero Data Retention. Ashby disables optional response storage; that setting does not eliminate abuse-monitoring retention. Provider account, billing, and operational metadata may have separate retention periods.

See OpenAI’s API data controls and Ashby’s subprocessor list. AI processing is not a promise that data remains in the Center’s country or Ashby’s database region; applicable DPA transfer safeguards continue to apply.

For Centers that accept this agreement, this section replaces only any earlier statement that the optional AI provider retains no content beyond inference. All other privacy protections and DPA obligations remain in force. It does not change how non-AI features process data, Ashby’s child-record retention policy, or the prohibition on selling data and using Customer Data for model training.

4. Review before sharing

AI drafts can be incomplete or inaccurate. Staff must check and edit each draft and deliberately approve sharing through the existing report or message workflow. Generating a draft never publishes a report or sends a message. Staff remain responsible for the final content.

5. Turning AI off

The owner may turn AI off in Settings at any time. New drafting requests are then blocked for the whole Center. Turning it off does not recall requests already transmitted to OpenAI, shorten retention that applies to earlier requests, or delete writing already saved in Ashby. Existing messages and reports remain subject to the Center’s normal record-retention rules. Reusable saved wording keeps the separate 90-day expiry described above and remains available with AI off until it expires or is deleted. To enable AI again, the owner must accept the then-current AI terms.

6. Changes and questions

Material changes to the AI provider, permitted data, retention, or these terms require an updated agreement and fresh owner acceptance before further use. Existing subprocessor-notice obligations continue to apply. Declining a new AI agreement leaves AI off without blocking other Ashby features.

Privacy questions: privacy@ashbycare.com.