Family far away, a child who does not want video left behind

When the viewing opportunity of distant family and the child's will are considered together, to whom and in what way should the performance be shown?

This page records a decision that has not been made yet. A group of people who are arranging a children's performance has to settle how that performance will be shown. Some families cannot travel to it. Some children would rather the performance left no video behind. A guardian may consent to recording while the child does not. And any video that leaves the group's hands — even one sent only to invited people — can be copied by a recipient and carried to a place the group can no longer reach.

Three options are on the table: livestream the performance publicly, provide video to invited people only, or keep the performance to on-site viewing. Nothing in this piece decides which option is right.

The spine runs in one direction: question, prediction, check against the outcome. First comes the question the meeting must answer, and the exact limit of what its authority reaches. Then comes the prediction: thirteen AI-generated judgments written on the same record. Then come the two event branches, drawn from the same starting point as alternatives. Then comes the check: the observations that must be fixed now, in writing, so that the decision can later be assessed rather than merely remembered.

1The scene in the decision year

Synthetic scene — the scene below is an invented illustration of the record's premises. It establishes no fact beyond what the record states.

It is the first meeting of the preparation group in 2026. Six people sit around a table in a community hall. There is a chair who keeps the agenda. There is an operations lead who has been reading the technical notes. There is a guardian who came in from the next district, and a child who has come with them and is sitting at the end of the table.

On the table is a list of the people who have asked to watch. Names, then towns, then notes in the margin: an aunt in another country, a grandparent who cannot travel, a cousin who works nights. Everyone on that list is hoping to see the performance from somewhere far away.

The chair opens a second folder. It holds one sheet per child. Beside most of the names, someone has written "consent: yes" in a guardian's handwriting. The chair reads one aloud and looks for a second answer. There isn't one.

The guardian speaks first. They are happy for the child to be filmed. They have signed the consent. They would like the recording to exist, because the aunt in another country will never otherwise see this.

Then the child says the thing the meeting has to hear. They are not refusing to perform. They will sing, they will bow, they will be there. They are refusing the idea that the performance gets left behind — that a copy of their face and their voice stays somewhere after the day, held by people they have not chosen. The guardian and the child are both answering honestly, and the two answers do not match.

The operations lead works through how the other option would work. A short clip, the lead says, sent to one invited household — and then forwarded, perhaps the same day, to a group chat. Held by people the group has never met, and no one in this room could pull it back.

The room goes quiet for a moment, because the technical problem and the moral problem have arrived together. If the group films, it may betray a child who said no. If it does not film, it takes something away from relatives who cannot come, and from any child who wants the performance to outlive the day. If it films only for invited people, it narrows the exposure without removing the copy risk.

Nobody in the room knows three things the record leaves unverified. Nobody has yet asked each child what they actually want about filming and disclosure. Nobody has found out whether distant family could be connected to the performance in some way that involves no video. Nobody can say how much would be recoverable if a child later changed their mind and asked for deletion.

The chair writes the question on the board and underlines it once:

The meeting has twenty-one days. By the time the viewing guide is printed, the scope of filming, transmission and video provision has to be settled — in a way this room can record for the children performing and their guardians, and explain to the child who said no.

2The question on the table

The question has a shape, a holder, and a date. It is not "should the performance be recorded?" That would be the easy question, and it is not the one the group faces. The group has to decide how wide the showing of this performance is to be — publicly livestreamed, given only to invited people, or kept to the room — and the record states the task in exactly those terms:

The decision actor is named, and it is a committee, not an individual. A preparation meeting is a temporary body: it exists to get a performance ready, and when the viewing guide is issued its work is largely done. That matters for how its choices land. Whatever it fixes is a decision about this performance, not a standing rule, and the judgments on the record are careful to keep saying so. A decision that outlives its performer is a different kind of decision from the one this group is making.

The mandate is short enough to memorise:

Three objects, and they do not all belong to the same owner. Performance video is an artefact the group can make or decline to make. Child's will is something the group can ask for but does not hold; it belongs to each child, and it can be given, withheld, or changed. Family access is a benefit that exists only if the group decides to extend it, and it is the thing that distant relatives are counting on.

2.1Who decides, and by when

The clock is already running when the meeting sits down, and the record fixes it without hedging:

The viewing guide is the document the group will hand to performers, guardians and viewers. It is the moment the scope becomes real: once it states whether there will be filming, whether there will be transmission, and who may see what, the question has been answered in practice whether or not anyone remembers the argument that preceded it. The reason the deadline bites so hard is given in one sentence, and that sentence carries both the pressure and the contradiction at once:

Two facts sit in that sentence, and neither of them is negotiable. Distant families exist. Children who want no video behind them exist. Any option the group adopts has to be defensible to both.

2.2Where the authority stops

A decision is only as useful as the boundary drawn around it, and the judgments on this record draw the boundary tightly. One of them states it plainly:

So the group's choice reaches its own filming, its own transmission, its own provision of video, and the notice it sends out about personal filming on the day. It does not reach the wider world of the hall. It does not decide what other people do with their phones. One judgment states that whether personal filming by the audience can actually be controlled is not answered anywhere in the record, and nothing else in the record settles it. It does not reach other performances, and it does not reach any outside organisation that might later want the material. One review states the limit from the side of what the meeting cannot promise:

This is the part of the question that makes it difficult rather than merely careful. The group can decide what it will do. It cannot decide what a recipient does with a forwarded file. The most a guide can do is state where access ends and what the group itself can delete; it cannot reach the copies.

2.3The three options, and what each costs

The first option is named in the record as follows:

Its trade-off is stated in one line of the record:

Widest viewing, widest exposure, and the least reversible of the three. Every judgment on the record rejects it, though they reject it for different emphases rather than in one chorus.

The second option, invitee-restricted video, carries a dedicated trade-off line in the record, stated in these exact terms:

It is the middle choice in width and the choice that still creates an artefact. Once a file exists and has been sent, the risk of copying and redistribution has not been eliminated by the act of sending it to a smaller list. The judgments that choose this option do not claim otherwise; they claim that the exposure can be narrowed, the child who refuses can be excluded from screen and voice, and the residual risk can be accepted knowingly rather than passed on to a child.

The third option, on-site viewing only, is described as reducing long-term exposure of video while forfeiting the sharing opportunity with family who cannot come and with the child. It is the only option that leaves no artefact of the group's own making. It is also the only option that answers distant relatives with nothing but the hope that some non-video connection will turn out to exist. Its cost is not a risk; it is a certainty. Someone will miss it, and the record does not soften that.

2.4What stays unknown until it is confirmed

Three things are not known, and the group is not allowed to behave as though they were known. The first is the one that governs the whole argument:

The second is the non-video connection for distant family already raised in chapter 1, which the judgments likewise treat as unverified. The third is how much would actually be deletable if a child later changed their will, which is a question about recoverable copies, about the group's own stored original, and about access that has already been granted.

None of these three may be filled in by assumption. One judgment states the rule the others follow: the filming and disclosure scope each child wants must be confirmed individually, with the child, and a guardian's answer cannot be substituted for it. Another adds that a non-response is not consent. A third adds that any proposed connection without video must not be treated as an alternative if it would still record the child's appearance or voice.

This is what makes the decision a decision rather than a preference. The group has to choose a scope now, in a state where the facts that would settle it are unknown and have been given people in charge and confirmation windows to be found out. It has to choose knowing that the choice may be reopened if the confirmations come back and change what the options cost. The question on the board, in the end, is not only how the performance should be shown. It is also who stands to lose something here, and whether the group can bear that loss before the confirmations arrive.

3The prediction: convergence and split

The record carries thirteen AI-generated judgments on the same question, all written on 2026-09-08, all at the same confidence. Before looking at the split, it is worth being exact about what the thirteen share, because the shared ground is larger than the disagreement — and it is shared for reasons none of them has to defend alone.

3.1Where the thirteen agree

The first agreement is negative and total. No judgment on the record chooses public livestream. Not one. Thirteen independent positions, each with its own named perspective, and a single unanimous rejection of O1, carried on the same three grounds that chapter 1 put on the board. Some children refuse video. Even a restricted copy is hard to control once a recipient has it. Unverified consent is not grounds for disclosure. Nothing in the record is contested about these three; the contest is about what follows from them.

The second agreement is about confidence. Every one of the thirteen says MEDIUM. That is not an average reached by negotiation — each judgment arrived at it separately, from the same premise: all three unknowns remain unverified. The judgments say repeatedly that the filming and disclosure scope each child wants has not been observed, that a non-video connection for distant family has not been demonstrated, and that the deletable scope when a will changes has not been measured. A MEDIUM confidence here is not timidity. It is the honest rating of a decision made on settings that are known to be settings, before the observations that would settle it have been obtained.

The third agreement is about scope, and it is unusually clean for a decision of this kind. Several judgments limit the effect of their own answer to the official video of this performance, the child's appearance and remarks, and the family's way of accessing it. They decline to extend it to future performances, to external organisations, or to any general child-filming policy. Two of them make the point from opposite directions: one states the limit in terms of what the decision covers; the other in terms of what it cannot promise anyone about the copies that already exist.

Two judgments were written with their Decision field naming the option in full rather than by letter — Decision Quality Review and Outcome Review each write "O3: On-site viewing only"; the other O3 judgments name the option by letter alone. The naming is a small thing, but it signals something about how the split is read. These two do not treat the choice as shorthand.

The fourth agreement, and the deepest, is that the decision is a loss allocation rather than a question of who is right. One piece of material in the record states it without hedging: the benefits and costs of the option accrue to different parties, and which loss to accept remains undecided. Both camps build their answer on this. Neither can claim the material as a verdict in its favour, because the material carries no verdict in it.

3.2Eight judgments: on-site viewing only

Eight of the thirteen choose O3. Their reasoning, taken together, runs as follows.

That sentence is the sharpest statement of the majority position, and it rewards slow reading: this is not a temporary hedge, it is a choice between two options, made and stated as made. The reason behind it is set out by another judgment, which frames the argument as a refusal to reason from something unverified:

Here the guardian's signature is present, but the answer a child gives directly to the person in charge is the answer that governs. A third judgment writes the same refusal as a limit on what may be assumed before the confirmations arrive:

Four other judgments reach the same choice with different emphases. Operations Planning frames it operationally: an unverified will is not consent, and a provisional decision taken under uncertainty is preferable to a scope that may have to be withdrawn later. Independent Challenge frames it as provisional in the strongest sense — explicitly not a permanent policy, covering this performance only, reopenable at the same scope before the guide is issued. Solo Judgment frames it as a question of reversibility, as chapter 2 already set it out. Outcome Review and Solo Judgment each add that under their own answer the performance is not to be cancelled — on-site viewing itself is permitted, personal filming on the day is managed to match the official scope, and a prohibition on personal recording that cannot be operated in substance would stop the recording operation or the guide that would allow it, rather than the performance.

What this camp accepts losing, it states openly:

The word "relinquished" is doing work. It is not "reduced" or "narrowed"; it is given up entirely, for this performance. Several of these judgments add that the loss may concentrate on specific people — a particular distant relative, a particular child who wanted the performance to outlive the day — and none hides this. Outcome Review says plainly that the loss is not small. Solo Judgment says the loss concentrates on family and on children who want to share, and adds that a connection method without video is to be confirmed separately but is not assumed to exist before confirmation. Judgment Synthesis, which is not in this camp either, makes the same point in the other direction: the loss of video sharing can fall on the distant family of a child who refused filming, which is the exact loss the majority has decided to accept.

Their conditions before acting are nearly identical in substance. The viewing guide must state that no official filming, livestream or provision will take place, and the person who confirms that this is so must be named. Each child must be asked, individually and separately from the guardian, whether they will be filmed and what disclosure scope they permit; a conflict between guardian and child must be recorded as a conflict, and the child's refusal must not be substituted for by the guardian's consent. Any proposed connection without video that would still record the child's appearance or voice is not an alternative, and may not be counted as one.

Their stop conditions are, in practice, a list of the things that would show the decision to be unworkable in the field rather than wrong in principle: official recording or transmission appearing during preparation; filming that overrides a child's non-disclosure will; the possibility that recording and redistribution cannot be controlled on site at all. In that last case one judgment goes further than any other on the record. If a guide of on-site only cannot itself hold — if recording and redistribution cannot be controlled even with no official recording — then the current plan is rejected, the related proceedings stop, and the scope is decided again before the guide is issued.

Their uncertainty statement is the one thing all eight share without variation. All three unknowns are unverified, so none claims that on-site viewing only is the option best for every child and family. What they claim is narrower: that under the current state of knowledge, avoiding irreversible recording exposure first is the safer default, and that this claim is worth MEDIUM confidence and no more.

3.3Five judgments: invitee-restricted video

Five of the thirteen choose O2. They do not treat the choice as unconditional, and the conditions are where the real disagreement with the majority sits.

That "however" is the hinge. This camp does not reject the majority's concerns; it attempts to meet them by narrowing the scope until they are met. Judgment Synthesis states a version that goes further than the others:

Under this formulation a child who refuses filming is not merely left out of a shot; they must not be identifiable in it at all, and their speech must not be recorded in it. Improvement Design sets the same threshold in its own terms, adding that a child whose will is unconfirmed is treated as refusing for the purpose of the exclusion test. Authority and Legal Review makes the internal operating principle explicit: where the guardian's will and the child's will conflict, the group's own rule is not to record that child. Decision-Maker's View adds that if even one child's appearance and voice cannot be reliably excluded, the group stops filming and stops provision altogether. On this side the switch to O3 is treated not as a failure but as the designed fallback — the argument is that O2 is defensible only while the exclusion test passes.

The accepted loss of this camp is stated with unusual candour:

None of the five claims that restriction removes the risk. They claim the risk is narrowed, that the group can state where access ends and what it can delete itself, that the copy risk is accepted knowingly by the group rather than transferred to a child who never agreed to it. They accept two further costs in the same breath: that the recording may be incomplete if a child refuses filming, and that some distant family will not be on the invitee list at all, and will therefore lose access a public livestream would have given them.

Their conditions are heavier than the majority's, which is itself an argument. Before any provision, each child must be confirmed individually on three points — whether filming takes place, what goes into the video, and what may be shown to invitees — recorded distinctly from the guardian's answer, with a non-response recorded rather than read as agreement. The invitee list, the authentication method, the access end point, the notice prohibiting redistribution, and the deletion scope the group itself can reach must all be documented before the guide is finalized. The point at which access ends must be set, and the group must be able to say which stored copies it can delete directly. If the deletable scope cannot be confirmed, the group must not announce that deletion is guaranteed.

The uncertainty section on this side is where the split is most honest. The five state that O2's feasibility is undetermined, and they name the signals that would change their answer. If many children turn out to refuse filming, the relative cost of O3 falls. If deletion does not reach the recipient's copy, the relative cost of O3 falls further. Conversely, if consent to filming and the restricted scope turn out to be clear and uncontested, and if no means of connecting distant family without video can be found, the grounds for maintaining O2 strengthen. Every one of these judgments says, in effect, that their choice is a prediction about what the confirmations will show — and that if the confirmations show otherwise, they will switch.

The strongest form of that switch is stated by Decision Quality Review:

Four conditions, and the fourth is the one that matters most: the parties accept the residual loss. That is not a technical finding. It is a judgement by the people to whom the residual risk would fall, and it cannot be produced by a test.

3.4Where the disagreement actually is

Reading the eight and the five side by side, the disagreement is not about the facts. Both camps accept the same three unknowns, both cite the same setting that even a restricted copy is difficult to control, and both reject public livestream outright. The disagreement is about two things.

The first is how to weight an irreversibility that has not yet happened. The majority treats the possible copy of a child who refused as the loss to avoid first, because if it occurs it cannot be undone, and because the person who bears it did not choose it. The minority treats the certain absence of family who cannot travel as the loss to avoid first, because it will occur, and because the group can reduce the other risk by narrowing the scope rather than by abolishing the artefact. Both sides are reasoning about losses that fall on different people, and both are doing so while the size of each loss is unknown.

The second is what the group is willing to risk in order to find out. The majority's answer is that the risk should not be taken at all until the confirmations are in. The minority's answer is that the risk can be taken under conditions that make it survivable, and that the confirmations should be completed as a condition of taking it rather than as a reason not to take it. Neither answer is a fact about this performance. Both are claims about how a committee should behave when its evidence is not yet in.

The record's value material declines to settle this, and so does this page. The question, already posed in chapter 2, is which loss the meeting accepts, and to whom.

3.5The timeline from decision to deadline

The thirteen judgments do not share a single schedule, but their milestones lay over one another without contradiction, because they are all observation plans for the same three unknowns and all expire at the same guide. The following is that combined timeline in relative time, assembled from the confirmations, persons in charge and confirmation windows that the judgments set out. It is an observation plan, not an observation result, and nothing on it has been confirmed.

Day 0the meeting fixes the scope for this performance; no official filming, livestream or video provision until the guide is finalized.

Within 7 daysthe designated person in charge confirms each child's filming and disclosure scope individually, and records any conflict with the guardian's opinion; non-responses are recorded, not treated as consent.

Within 12 to 14 daysthe operations lead tests a non-video connection for distant family and documents the scope deletable if a will changes; until then, neither is presumed feasible.

Day 18the meeting compares the confirmation results against the accepted loss of O3 and the redistribution risk of O2.

Day 21the viewing guide states the scope and the filming rules; if the cost of an option has changed, the decision is reopened rather than finalized.

Two differences are worth noting in the schedules, because they track the split rather than the deadline. The judgments that choose O2 test the deletion scope a little later, in the range of days 12 to 16, because their decision depends on that scope being documented before provision begins — one sets day 16, another documents the scope before the guide is finalized and checks again after the performance. The judgments that choose O3 tend to use two pre-guide confirmation windows instead: one for the child's will and the non-video alternative, and a second, immediately before the guide is finalized, for the deletable scope and the limits of copy prevention. The distinction is not cosmetic. The O2 camp needs the deletable scope as a condition of acting; the O3 camp uses it as the criterion for re-judgment, and does not act on the video question until it has been obtained.

One further point belongs on this timeline and has no day attached to it. Multi-Perspective Final Judgment states the reopening condition in its shortest form:

The phrase to hold onto is "loss structure". A confirmation that merely fills in an unknown is not, by itself, a reason to re-decide. What reopens the decision is a confirmation that changes what the options cost — who gains, who loses, and how much. That distinction is what the check in the next chapter is built around.

4The branches: a refusal on the day, and a repost

The record carries two event cards. They are not a sequence, and they are not a schedule. Both diverge from the same initial state — the decision taken before the viewing guide, in a state where all three unknowns are still unverified — and only one of them may occur. Read together, they answer a question the thirteen judgments all raise but none can settle in advance: what happens when one of the two premises the decision was built on turns out to be live rather than theoretical.

The first card is about the child in the room, the guardian's signature, and the gap between them. The second is about the file, the recipient, and the place it goes after it leaves the group's hands.

4.1Branch one: a child refuses on-the-day filming, contrary to the guardian's consent

The word that carries the weight is "on-the-day". This is not a child who refused weeks earlier, in a conversation the group could have acted on. It is a child who refused at the moment of performance, after the viewing guide was drafted, after the guardian signed, after the group had built its plan around the answers it expected to receive. The refusal arrives late, when the cost of honouring it is highest, and when the cost of ignoring it is also highest.

What happens next depends on which answer the group has adopted — and here the two camps diverge sharply, not because one of them would override the child and the other would not. None of the thirteen says the guardian's consent can substitute for the child's will. The divergence is about what the refusal proves.

For the camp that chose on-site viewing only, the refusal is a confirmation of the premise they were already acting on. One of the strongest statements of this position comes from Evidence and Sources Review, which sets its rejection condition in a form that covers this exact event:

In the on-the-day scenario, the question the group faces is not whether official recording begins — under an on-site-only guide, none should. The question is whether the child's refusal can be honoured in practice at all. If the group is operating on-site only, the refusal is a fact to be handled inside the plan: no official camera, no official transmission, and a managed approach to personal filming on the day. If the group is operating under an invitee-restricted plan, the refusal is a test of the plan's central condition.

That test is stated by Authority and Legal Review, and it is the most precise stop condition on the record for this branch:

Three things must be absent from the restricted video — screen, voice, and name — and the word "substantially" is doing real work. A child cannot be merely left out of the principal shots and left audible in the background. The exclusion test has to hold in the recording itself, not only in the plan. Improvement Design's treatment of an unconfirmed will as a refusal, set out in chapter 3, carries the same effect here: the burden of the ambiguity falls on the recording, not on the child. Decision-Maker's View states the threshold in its bluntest form, set out in chapter 3, and reaches the same result here. Domain Analysis reaches the same outcome from a different direction — if a child refuses filming, that scene or the entire performance is not provided as video, because the video would be incomplete in a way the group cannot fix afterwards.

There is also a stop condition that belongs to neither camp alone, and it is the one that treats this branch as evidence rather than as an inconvenience. Review Coordination sets it as follows:

The on-the-day refusal is precisely the kind of event that produces such a requirement. If a guardian insists that filming must go ahead despite the child's refusal, and if the group concludes that the performance cannot be delivered under an on-site-only plan that does not permit it, then the plan itself has been shown to be unworkable, and the scope has to be decided again rather than patched. Operations Planning sets a similar condition in operational terms: if the operating plan includes official filming, transmission or video provision while the child's will or the deletable scope is unverified, finalization of the guide stops. Decision Quality Review goes further and treats the failure as a rejection condition for the whole operating plan, in the terms set out in chapter 3.

The branch also carries a possibility in the other direction, and it is the one the on-site-only camp has to answer for. If the refusal is one of several — if a number of children refuse, or if no child refuses and all agree to a restricted scope — then the premise that blocked review of the invitee-restricted option has weakened. Independent Challenge states the condition for re-judging in exactly these terms:

This is the branch's real function on the record. The on-the-day refusal is not a reason to cancel the performance — several judgments are explicit that on-site viewing continues — and it is not, by itself, a verdict on either option. It is a signal that the scope must be re-decided at the moment when the guide can still be changed, and that the group must record what the refusal shows about the filming rules rather than merely handle the disruption.

4.2Branch two: part of the invited video reposted on a public account

The second branch begins after provision, and it begins with the file already outside the group's control.

The record states the event in one sentence, and it is the event that Premise P3 predicted without predicting when:

This is the scenario the invitee-restricted option is designed to survive and cannot guarantee to survive. The record does not confirm that the group sent its file to a list it chose, with a notice prohibiting redistribution, and with a stated access end point; those are conditions the O2 judgments require before provision, and the record does not show they were in place when the repost occurred. A copy reached a public account, and how it got there is not stated in the record. Nothing about the group's care in choosing the list changed that outcome, because the group's authority over the copies was never real — only its authority over the original and over its own stored copy.

The stop conditions on this branch are the sharpest on the record, and they are written to fire immediately rather than after investigation. Judgment Synthesis states the first and most direct response:

Two things happen at once: new access stops, and the deletable scope is verified — not assumed, not promised, verified. Improvement Design adds the consequence in its own terms: if the promised change-of-will and deletion procedure cannot actually be carried out, new provision and remaining access are stopped, and O2 becomes a candidate for discarding. In that case, if controlled restricted provision does not hold, the decision goes back to O3. Improvement Design states the alternative path as well — if copying or redistribution beyond the invited scope is confirmed, the group does not wait for a second signal before narrowing the plan.

Authority and Legal Review sets a further condition that this branch puts directly to the test: if it is verified that copies already created cannot be recovered, further provision stops and the scope of harm and whether to resume are decided again. That is a different task from verifying the deletable scope, and the record treats it as different. Verifying what the group can delete tells the meeting what it can still reach. Verifying what cannot be recovered tells the meeting what it has lost the ability to affect. Only the second of these tells the group whether the option it chose is still defensible.

Decision Quality Review states a rejection condition in similar terms, and it is worth noting because it applies the same test the on-site-only camp applies to itself: if it is verified that official recording or transmission is scheduled or is actually operating, the creation and transmission of that video must be stopped immediately. In the reposted-branch scenario, the relevant verification is not that official recording is operating — it is that the group's own provision has produced an artefact that has left the scope it was given. Outcome Review adds a signal it plans to check after the performance: whether leakage or a withdrawal request has occurred, and the scope of possible response. If the scope of possible response proves narrower than provision assumed, the deletable scope must be verified before any further statement is made to the invited list about what the group can do.

Solo Judgment does not set its stop condition to the repost itself. Its condition runs from the angle of the performance: if official or personal filming or transmission begins, or if a child newly states or withdraws a will not to be exposed, the stoppage of that filming and sharing procedure occurs immediately. Outcome Review adds a condition specific to the guide's own rules: if the prohibition on personal filming cannot be operated in substance, or if a performance video is already being generated and transmitted, the issuance of the viewing guide stops, or the filming operation stops, and the control measures are set again. Neither of these is the repost itself, but both describe the state the repost reveals — a guide whose rules have been overtaken by an event outside the guide's reach.

What makes branch two different from branch one is that branch one tests whether the plan can hold before the artefact exists, and branch two tests whether the plan can survive once it has been made. In branch one, the group still has the option of not filming at all. In branch two, the group has already filmed, already sent, and already lost some control over what it sent. The question in branch two is not whether to create the exposure but what remains of the group's ability to reduce it, and how much of the guide's promise about deletion was ever true.

4.3What both branches share

Neither branch resolves the decision. Both branches re-open it, and both do so under conditions the judgments have already written down.

The common shape is this. Each branch arrives from outside the confirmations that the 21-day window was built around. In branch one, the confirmation of each child's filming and disclosure scope has been overtaken by an event — a refusal at the moment of performance, made against a guardian's consent, at a point where the plan can still be changed but only just. In branch two, the confirmation of the deletable scope has been overtaken by an event — a repost on a public account, at a point where the group's authority over the copies no longer exists and never did.

Both branches therefore push the meeting toward the same three moves: stop the thing that is causing the exposure, verify the scope that is actually left, and decide again whether the option the group chose is still defensible on the evidence that has just arrived. And in both branches, the verification is what matters, not the alarm. A repost does not, on its own, prove that invitee restriction was always unworkable — it proves only what the record states — that part of the invited video was posted again on a public account — and the group has to weigh that alongside the deletable scope and the strength of the notice. A refusal does not, on its own, prove that the on-site-only plan was right — it proves that at least one child's will and the guardian's signed consent were not the same thing, which the record has held as a setting all along and now has to hold as an event.

There is one point the judgments are explicit about, and it belongs at the end of this chapter rather than buried in it. Neither branch resolves in the direction the group might hope. The record is careful to state that the signals arising from these branches do not settle the question by themselves. Evidence and Sources Review puts it plainly:

That sentence is not a hedge. It is a rule about what the branches can and cannot do. A verified repost does not automatically prove that invitee restriction was the wrong choice, because the group's answer to a repost depends on what the deletable scope turns out to be — and that scope has been unverified from the beginning. An on-the-day refusal does not automatically prove that on-site viewing only was the right choice, because the group's answer to a refusal depends on whether the filming rules it wrote can be operated in practice. And in neither branch does the event revive public livestream. Every judgment on the record declines public livestream, though not all of them say so in the same words, and no branch produces evidence that would revive it. One judgment says it in the form this chapter ends on: even where all the re-examination conditions are met, there is no automatic switch to public livestream.

The branches are alternatives, not a sequence. They diverge from the same state, they fire different stop conditions, and they leave the group with different tasks. What they share is the thing the whole record has been building toward: a decision made in advance of the facts that would settle it, with the reopening conditions written down now, so that when one of them arrives, the group can act on what it said it would do rather than on what the moment makes it want to do.

5The check: what will be observed

A decision made before the facts that would settle it is only defensible if the observations are fixed in advance. Otherwise the outcome is assessed against whatever the moment happens to make salient, and the group gets to redefine, after the fact, the loss it was accepting. The point of this chapter is to settle that question now, while the deadline is still ahead of the meeting and the confirmations have still not been obtained.

The table below is the check. Its first column states the observation. Its second states who will make it and when, so that the observation cannot quietly be skipped or quietly be redefined. Its third states which judgments the observation would bear out or weaken — not which option it would prove, because no single observation in the record proves an option. Its fourth states what the observation reopens: the part of the decision that would have to be argued again if the observation comes back one way rather than another.

The state of the record as this chapter is written is plain: Before the deadline; awaiting the check.

The first row is the one that governs the other six. Almost every judgment on the record treats the individual confirmation as a precondition for acting; one judgment states explicitly that, for on-site viewing only, the confirmation is not a condition delaying implementation but a condition for re-judgment. For the on-site-only camp, the confirmation is what makes the provisional decision defensible, because it establishes that the group asked each child and did not read a guardian's signature as the answer. For the invitee-restricted camp, the confirmation is what makes the option executable at all, because the exclusion test cannot be run against an unknown will. Neither camp is asking for a fact the record already holds. Both are asking for a fact nobody has yet obtained.

The second and third rows are the ones the camps weigh differently. The majority's accepted loss is the access that distant family lose; the minority's accepted loss is the copy risk that survives restriction. Row two measures the size of the first loss — whether some non-video connection exists that would reduce it — and row three measures the size of the second, whether the group can actually reach what it has given away. Together they are the two halves of the comparison the meeting has to make on day 18, and the two halves that neither camp can currently fill in.

The remaining four rows are the ones that come from outside the plan. Three of them are branch observations rather than confirmation windows: a detection of official recording during preparation, a repost of the invited video, and a refusal on the day. These are not on the confirmation schedule, because they are not things the meeting obtains — they are things that arrive. The judgments have written their stop conditions in advance precisely so that these arrivals produce a defined response rather than an improvised one. The fourth row in that group, the guide's own statement of scope, is the check on the decision itself: it is the point at which the meeting's answer becomes a document that others will act on.

5.1The three re-judgment triggers

The record does not merely ask for observations. It asks that each observation be tied to an object, a person and a window, and that the decision be reopened only if the observation changes what the options cost. That requirement is stated three times, once for each unknown, in language that has not been softened:

Read together, the three carry a rule the judgments state in several places. A confirmation that merely fills in an unknown does not, on its own, re-decide anything. As chapter 3 set out, what reopens the decision is a confirmation that changes the loss structure. A finding that the specific will of every child concerned permits filming, and that the actual scope that can be deleted upon withdrawal is confirmed, weakens the premise that had blocked review of the invitee-restricted option. A finding that distant family can be connected without video reduces the loss the on-site-only camp accepts in writing. A finding that deletion is far narrower than any guide could honestly describe removes the ground on which restricted provision was to be justified at all. Each of these changes a cost, and each is therefore a reason to argue again rather than a reason to be proved right.

5.2What the check will not do

This chapter fixes observations. It does not score them. The table carries no verdict column, because the record supplies none: the piece of material governing the whole question says that the benefits and costs of the option accrue to different parties, and that which loss to accept remains undecided. That is not a gap to be closed by a better table. It is the thing the meeting decides.

The present state is therefore a state of waiting, and it should be named as one. The confirmations have not been performed. The schedule above is an observation plan, not an observation result. Three of the seven rows describe events that may never occur, and a fourth describes an outcome of a procedure that has not yet been run. If the deadline passes with the confirmations completed and the observations recorded, the group will be able to judge the decision against the observations it fixed in advance. If it passes with the confirmations not completed, the group will have fixed a scope on the strength of assumptions it was explicitly required not to make — and the check will record that too, which is the point of writing it down now rather than afterwards.

6The questions that remain

The record carries three questions that were not written to settle the decision and were not resolved by any of the thirteen judgments. They are recorded as watch questions, and they are worth keeping beside the reopening conditions for the same reason: the answer changes what you look for when the confirmations come back.

The first is a question about value, and it is asked in the passive, which is how the record keeps it open.

One reading — offered here rather than by the record — is that the value of the experience has several contributors. The children who performed created part of it. The guardians who signed created part of it. The family who cannot come because they are far away created part of it — the record lists their far-away situation among the reasons this decision is being made, and under on-site-only viewing they lose the chance to watch. None of these parties created it alone, and how the meeting distributes the viewing opportunity among them is a question this reading raises rather than settles.

The second question turns on what fairness means when the participants are not the same.

This piece offers a parallel the record does not settle: One set of conditions would be to ask every child the same question and treat the answers alike — including a guardian's signature and a child's refusal, which are not the same kind of answer. One set of opportunities would be to give each child what each child asked for, even when the answers differ, and to carry the resulting complexity in the guide rather than smoothing it away. The meeting does not have to choose between the two definitions to notice that it is choosing between them: a rule that treats consent uniformly and a rule that treats each will individually will produce different videos, different invitee lists, and different losses.

The third question is about where the obligation stops.

In this record the boundary has been drawn several times and never reached a clear end. The meeting's authority reaches its own filming, its own transmission and its own stored copy. It does not reach a recipient's forwarded file. It does not reach a public account that receives that file without ever having been on the invitee list. And it has not been established whether it reaches the family members who are far away and cannot come, and whose access the record treats as a benefit to be extended rather than a right to be met. In this piece's reading, the obligation question in this record is concrete rather than abstract: whether the group owes distant family a viewing, whether it owes each child protection from exposure, and whether it owes both — a question the record leaves open when it says only that which loss to accept remains undecided.

Each of the three questions can be asked in your own organisation or community, and the third is where most groups find the decision already half-made by habit. Where does your obligation to participants extend: to the moment of delivery, to the artefact it produces, to the copies of that artefact you can no longer reach, and to the people who were never in the room and had no say in the rule? In your own organisation or community, a parallel decision may already exist — a recording, a listing, a shared file, an archive kept "just in case" — and you might ask who created the value in it, who gets the same conditions and who merely gets the same opportunity, and how far past the group's own reach the obligation is being drawn.

The record does not answer these questions, and neither does this piece. No option on the table is declared right here, and the table in the previous chapter carries no score, because the material governing this case says the benefits and costs of the option fall on different parties, and that which loss to accept is undecided. That verdict belongs to the person who signs the viewing guide. What this piece can do is keep the three watch questions beside the confirmations, so that when the person in charge reports what each child wants, what a non-video connection would do, and what deletion actually reaches, the meeting has the questions already open.

The first question is the one to ask before the others. If the contributors to the value of the experience are named, then the losses on the table are named with them, and the decision stops being a choice between three technical options and becomes a decision about which people this performance was for.

7Sources and notices

Two background documents stand behind the record, and both are cited as context rather than as evidence. The first is "UNESCO · AI and Culture expert report overview," published 2025. Its scope is the relationship between AI and creation, cultural heritage, and diversity; it does not adjudicate the truth of any particular faith or the ranking of future cultures. The second is "UNESCO · Ethics of Neurotechnology," published 2025, which sets out ethical principles touching mental privacy, autonomy, and identity. It is not evidence that memory duplication or consciousness uploading is possible. The record's note for the second document says the same as the first: it is background for the theme and the institutional questions, and does not prove the future events, figures, or technological realization of the case. Read together, the two point at the shape of the question rather than at its answer — who is created for, who is represented, and where a mental life may be touched without consent.

The record also carries a symbolic frame, chosen by its creator and offered as exploratory rather than as proof:

The thirteen judgments on this page use that frame as a way of finding the question about the children's will, not as a claim about how the case must be resolved.

The piece of material that governs the whole decision is the shortest of the four, and it is worth repeating here because it is the reason the check chapter carries no score:

The notices. Every person, institution, event and figure in this record is a synthetic setting, as chapter 1 sets out. Nothing on this page is a statement about an actual case or about current law. The thirteen judgments are AI-generated, written 2026-09-08 on the same input, and this piece is AI's expansion of the record: it quotes the judgments, sets them beside one another, and adds no verdict of its own. The record's own premise policy holds for everything above — source documents are the background and the roots of the norms, and they do not prove the premises of the case.

The input hash for the record behind this piece is b7f5d669b2cd618d800086148f2624e6c98bfb4414fda14053bf21eeaaa7545f. The record itself is an English translation of the Korean original.