Review Prompt to Map Your Exposure Under Knowledge Standards for Minors
This Luna review prompt audits your platform against the knowledge standards in current minor-protection law, from the state social media statutes in force and the California bracket regime to the federal KOSA/KIDS drafts whose "knows or should have known" logic is the direction of travel. Luna, the deepidv compliance agent, builds a knowledge inventory of every signal you already hold that implies a user's age and the share of active users each signal plausibly marks as under 13, under 16, and under 18, then draws an obligation map per jurisdiction showing which current defaults would violate the duties that attach once that knowledge is imputed. It ranks a gap register of surfaces where minors are plausibly present and no age assurance runs, per-minor penalty regimes first, then lays out a layered program design (estimation with liveness as the default gate, credential and parent-mediated escalation, appeals) and the evidence schema that answers an AG inquiry or audit. Built for trust and safety and compliance leads whose own data already implies which users are minors, and who can no longer treat "we don't verify age" as a defense.
How to use this prompt
- 1
Open Luna in the deepidv dashboard and paste the full prompt, or run it in Claude, ChatGPT, or Gemini if you are drafting the exposure map outside the platform.
- 2
Add an INPUT block below the prompt with the jurisdictions you operate in, the age signals you already hold (declared birthdates, graph clustering, content and session patterns, device signals), and the age assurance running on each surface today.
- 3
Run the prompt and read the knowledge inventory first: it shows the share of active users your own signals already mark as under 13, under 16, and under 18, which is the knowledge a regulator can impute to you.
- 4
Work the gap register top down, since it is ranked by penalty exposure, per-minor regimes first, and then user volume, and assign an owner and remediation date to every gap and every default the obligation map flags.
- 5
Re-run quarterly and whenever a state statute takes effect or the federal KOSA/KIDS drafts move, so the obligation map tracks the law as it changes.
The prompt
Luna, audit our platform against the knowledge standards in current minor-protection law: the state social media statutes in force, the California bracket regime, and the federal KOSA/KIDS drafts whose "knows or should have known" logic is the direction of travel. Produce: 1. Knowledge inventory: every signal we already hold that implies a user's age, declared birthdates and their contradictions, graph clustering, content and session patterns, device signals, and the share of active users each signal plausibly marks as under 13, under 16, and under 18. 2. Obligation map: per jurisdiction we operate in, which duties attach to users we know or should know are minors, feature restrictions, consent requirements, data limits, reporting, and which of our current defaults would violate them if knowledge is imputed from the signals in item 1. 3. Gap register: surfaces where minors are plausibly present and no age assurance runs, ranked by penalty exposure (per-minor regimes first) and user volume. 4. Program design: the layered response per gap, signals consumed, estimation with liveness as the default gate, credential and parent-mediated escalation, appeals, with data-minimization rails for the statutes that mandate deletion. 5. Evidence schema: the per-decision record and periodic reporting format that answers an AG inquiry or audit, detection coverage, method volumes, both error rates, appeal resolution times. Treat "we don't verify age" as a liability statement, not a defense, and date every remediation with an owner.
Test it in Claude or another LLM
This prompt is built for Luna inside deepidv, where Luna works from your live verification records and per-decision evidence. You can dry-run the exposure map in any general LLM first with synthetic signals and surfaces before pointing it at real users.
- 1
Paste the full prompt into Claude, ChatGPT, or Gemini, replacing the opening 'Luna,' with a role instruction such as 'Act as a minor-protection compliance auditor mapping a platform against knowledge standards.' Keep the five numbered deliverables as written.
- 2
Below the prompt, paste the synthetic sample block from this page so the model has jurisdictions, age signals, and surfaces to audit.
- 3
Add one framing line: 'This is synthetic test data. Where a statute's duty or penalty is not in the input, flag it as an open question rather than inventing it.'
- 4
Check the output shape: a knowledge inventory, an obligation map, a gap register, a program design, and an evidence schema, with an owner and date on every remediation. If any duty cites a statute the input did not name, tighten the framing and re-run.
- 5
Once the shape is right, run it live in the deepidv dashboard where Luna works from your real verification records.
Synthetic sample data to paste alongside the prompt
Fake test data, safe to share with any LLM. Swap in your own once the output looks right.
JURISDICTIONS (synthetic, fake): State A social media statute in force with per-minor penalties; California bracket regime; federal KOSA/KIDS drafts pending AGE SIGNALS (fake): - declared birthdates: 4% of active users declared 18+ but post school-year content - graph clustering: 11% of users sit in friend clusters where most members declared under 16 - device signals: 7% of sessions come from devices reporting the under-13 OS bracket SURFACES (fake): - signup: declared birthdate only, no age assurance - direct messaging: open to all ages, no age assurance - livestream gifting: facial age estimation with liveness, buffer to 21 OPEN ITEM (fake): whether State A's deletion mandate covers estimation selfies after a pass
Pairs with on deepidv
FAQ
What does a "knows or should have known" standard mean for platforms?
It means minor-protection duties attach not only to users a platform actually knows are minors, but to users its own signals should have identified as minors. The federal KOSA and KIDS drafts disclaim any age verification mandate, yet protections that attach to minors require knowing who the minors are, so a platform that holds age signals and never checks age carries the liability anyway.
What does the Minor Knowledge Standard Mapper produce?
A knowledge inventory of the age signals you already hold, an obligation map of the duties that attach per jurisdiction, a gap register ranked by penalty exposure and user volume, a layered age assurance program design, and an evidence schema built to answer an AG inquiry or audit. Every remediation comes back dated and assigned to an owner.
Can I run this outside the deepidv dashboard?
Yes. The structure works in Claude, ChatGPT, or Gemini as an audit framework and returns the inventory, obligation map, and gap register from data you paste. Working from your live verification records and per-decision evidence only happens when it runs inside the deepidv dashboard through Luna.
Related prompts
Run it with live verification data
These prompts work in any LLM. Inside the deepidv dashboard, Luna, Arbiter, and Arc run them against your real sessions, screening lists, and audit trails.
Book a Demo