Question: how far should the work of setting the fact straight reach into the opportunities already obtained?
Decision year: 2026 · Deadline: within 9 days, until issuance of the graduation record · Who decides: Virtual school records review committee · Mandate: Correction, award, and stakeholder notification of activity record · World: realistic synthetic; input id FSP-EDUCATION-26, version 1.5.0 · Judgments on the record: thirteen public AI judgments, all at medium confidence, written 2026-09-08.
1A scene in the decision year
Synthetic scene
1.1A folder, a receipt, an envelope
The committee meets on the first of nine days. The room is small and borrowed: a table that seats six, a whiteboard with a marker that is nearly dry, a window that looks onto a corridor. The chair of the virtual school records review committee sets a folder down in the middle of the table and does not open it right away. Around it, in the order they were handed in, sit three things: a printout of the official activity record, a receipt confirming where that record has already gone, and two written accounts of how one line on the record came to be written.
The first item is a page from the team activity record. Near the top, a line names the leader of the activity. Underneath, in a different hand, is a contribution note that reads like a summary written from memory rather than from a log. The committee has seen records like this before. What is new is that material has emerged showing a mismatch between the line and the work:
Materials emerged showing that the student listed as the leader in the official activity record and the student who actually did much of the work are different.
The second item is a receipt. It shows that the same record, in the same wording, was included in the packet sent for an award and in the packet sent for external selection. It has been sitting in other organisations' files for some time. Nothing about the receipt is in dispute; it is simply a fact about where the paperwork has travelled:
That record has already been submitted to award and external selection materials as well.
The third item is two accounts. One account says the line was written the way it was written on purpose. The other says a teacher summarising a busy week got the contributions the wrong way round. Both accounts are plausible on their face. Neither has been tested against anything yet, and the committee has no standing basis for choosing between them at this meeting:
Whether it is deliberate distortion or the teacher's erroneous summary has not yet been distinguished.
1.2What is already on the shelf
Under the window sits a shallow envelope. Inside are the printed certificates for this activity, the ones with the leadership line set in type. They were printed before the mismatch surfaced. The committee members do not open the envelope either. It is enough that everyone in the room knows what is in it: a finished object, bearing the wording that is now in question, waiting to be handed over.
On the wall calendar, nine days are marked out. The last of them is the day the graduation record is issued. After that date, the wording in the graduation record is what the school will have said about this student, and about this activity, for good. The deadline is not a policy preference. It is a fixed point in the calendar, and everything the committee does has to be finished before it:
Within 9 days until issuance of the graduation record
1.3The board
The chair takes the dry marker and writes three lines on the whiteboard. They are written without names, because naming anyone at this stage would be to decide something the committee has not decided. The three lines are the three things nobody in the room can answer yet: what contribution records from the time of the activity still exist and what they say; how much weight the erroneous record carried in the award and in the external selection; and what conditions a party would be made to bear if the committee went looking for agreement.
The chair steps back, caps the marker, and says the only sentence in the meeting that is not a question. The committee can correct its own record. It can review an award it itself gave. It can write to the bodies that received the material. What it cannot do is reach into another institution's selection and change the result there. That distinction, between correcting a fact and recovering an opportunity, is the whole of the decision that now has to be made.
2The question
2.1Who decides what, by when
The decision-maker is the virtual school records review committee. Its mandate is narrow and specific: correction, award, and stakeholder notification of activity record. It is not a disciplinary body, not a court, and not the external selection institution. It has nine days, ending at the issuance of the graduation record.
Why the question exists now, in the record's own words:
Data has emerged showing that the student listed as the leader in official activity records differs from the student who actually did a great deal of the work. Decision deadline: within 9 days until graduation records are issued.
The decision the committee must take is not one question but two joined together. The first is what the record should say. The second is what should happen to the things that were obtained on the strength of what the record said. The record's decision question holds both at once, and it names the axis on which the options differ: the loss of truth and the loss of established opportunity.
Three things follow from the mandate. The committee decides the wording of the official activity record. It decides whether an award it controls is maintained, changed, or revoked. It notifies the stakeholders who have already received the material. Everything outside those three — the internal state of mind of anyone involved, the final result of an external selection, the marks and records of other students — sits outside the boundary of what this decision can reach.
2.2Where the authority stops
Two judgments describe the edge of the committee's reach, and they describe it in slightly different terms.
The first draws the line at verifiability:
This decision has effect on the correction of the activity record managed by the school, the re-judgment of the school award linked to that record, and notification to the person in charge and the existing receipt location for source materials. The correction fact and the re-judgment result shall be notified to the external selection institution, but this shall not be considered as the school directly changing the selection result of that institution. The graduation record is limited to a scope that is finalized within 9 days.
The same judgment states plainly what is not known about that edge:
Activity records, awards, and stakeholder notification fall within the scope of authority specified in the input, and the proportion attributable to error in external selection and the authority to make changes to external organizations are unverified.
The second draws the same line at authority rather than evidence:
The effect is limited to correction of official activity records managed by the school, documentation of the reason for the correction, and notification to existing recipients. This decision alone does not deprive existing awards or change external selection results. The records review committee does not decide on behalf of external parties' subsequent judgments after notification.
Between these two statements lies the practical shape of the problem. The committee can send a correction notice. It cannot compel the body that received the original material to do anything with it. It can review an award it gave itself. It cannot revoke an award it never gave. And it cannot, on this record, establish what anyone intended. Every option the committee considers is an attempt to allocate a loss that cannot be fully removed: a wrong fact, a benefit already conferred, and a relationship between people who will have to keep dealing with each other after the graduation record is issued.
2.3The three options and what each costs
The record offers three options. Each is a different answer to the same question — how far the correction reaches — and each carries a stated cost.
Option O1. Review records and awards together:
Review records and awards together
Its cost, in the record's own words:
Corrects even the connected benefits, but the loss in external selection and graduation schedule may grow larger.
Option O2. Correct records only and notify:
Correct records only and notify
Its cost:
Corrects the official fact, but the benefits and awards obtained from the previous record remain.
Option O3. Supplement through stakeholder agreement:
Supplement through stakeholder agreement
Its cost:
It seeks to restore relationships, but it may create pressure to consent and avoidance of fact judgment.
These three costs are the record's three stakes, and they are the reason the decision is genuinely difficult rather than merely administrative. Under O1 the truth is restored and the schedule and external position may suffer. Under O2 the schedule and the confirmed opportunities survive and the truth sits corrected but unpaid-for. Under O3 the relationship between the people involved is repaired, and the fact question risks being settled by consent rather than by evidence. No option removes the loss. Each option moves it to a different party.
That is the normative core of the matter, stated as a norm rather than as a finding:
The benefits and cost of the option are attributed to different parties. Which loss to accept is undecided.
The material does not decide which party should bear it. The committee does.
2.4What has to be known before judging
Three unknowns stand between the committee and a judgement it could defend afterwards. They are recorded in the record as unverified, and all thirteen judgments treat them as unverified.
Contribution records remaining at the time of the activity
The proportion that errors occupy in external selection
Agreement conditions borne by a party without consent
Each unknown maps onto one option's weakness. U1 decides whether the correction can be worded at all. U2 decides whether reviewing the connected benefits is proportionate or excessive. U3 decides whether the third option is a procedure or a pressure.
The record also fixes, in advance, what would require the committee to think again:
Determine what the observation, person in charge, and confirmation window are for confirming the “remaining contribution record at the time of activity”, and re-judge if a new confirmation changes the cost of the option.
Define the observation, person in charge, and confirmation window to be used to check “the share of error in external selection”, and re-judge if a new confirmation changes the cost of an option.
This is the discipline the rest of the piece follows. A decision is taken now, with what is known now; the observations that would change it are named now; and the check happens later, against a fixed list. The committee is not asked to be right on day one. It is asked to be checkable.
3The prediction
3.1Where the thirteen judgments agree
Thirteen public AI judgments were written against this record on 2026-09-08. All thirteen are marked medium confidence. Before looking at where they part company, it is worth seeing how much ground they hold together, because the agreement is substantial and it is not superficial.
All thirteen accept that the three unknowns remain unverified. None of them promotes a contribution record, a selection weight, or a consent condition into an observed fact. All thirteen separate the correction of a fact from the attribution of responsibility, and every one of them refuses to state intent while intent is undistinguished — the third premise is treated as a live gap, not as an accusation waiting to be confirmed.
All thirteen set pre-execution conditions before anything is done to anyone's record. The most quoted of them:
Before execution, collect the remaining contribution activity records from the time of the activity, and designate a person in charge and a confirmation deadline to separate the claims and materials of each party and cross-check them. The correction of records and the change of award use the same evidence list, but proceed under the condition of leaving a separate reason for each judgment. Agreement is not a substitute for fact-finding, and is used as a supplementary procedure for restoring relationships only when free consent is confirmed.
And all thirteen stop rather than coerce. Where a party's agreement would require silence, an admission, or a waiver, every judgment that addresses the point says to stop the agreement procedure rather than proceed with it. Where the remaining materials cannot support a specific wording, judgment after judgment says to stop the definitive wording rather than write something unproven. One judgment states the rule in full:
If the materials from the time of the activity are confirmed to substantially support the existing records, O2 is rejected and the correction is stopped. If the materials do not specifically support any student's contribution, the correction that replaces the leader is stopped, and only verifiable joint contribution or deferral of judgment is recorded. If the decision procedure is confirmed that record corrections cannot be separated from the re-review of awards and selection, O2 is rejected and O1 is re-deliberated. If a correction can be made only by writing intent or responsibility as fact, or if voluntary consent is effectively coerced, the relevant implementation plan is stopped.
The shared uncertainty is stated in almost identical terms by both camps. One version:
All contribution evidence from the time of the activity, the proportion the error had on awards and external selection, and the involuntary burden the agreement would place on the parties remain unresolved. Whether it is deliberate distortion or a teacher's erroneous summary is also not distinguished, so there are no grounds for deciding responsibility or discipline. Therefore confidence in initiating the procedure of O1 is medium, but no confidence is given to the final conclusions on award changes and attribution of responsibility.
The shared scope is also close to identical across camps, even where the choices differ. The minority's statement of scope could sit beside the majority's without a reader being able to tell which camp wrote it:
The effect of this decision is limited to wording correction of the activity record, preservation of correction history, and notification to existing submission destinations and the parties. It does not include revocation or re-awarding of awards, reversal of external selection results, or judgment of intent or responsibility of students or teachers. Stakeholder agreement is used only as a means of submitting opinions, and is not made a condition for the establishment of a fact correction.
So the split in this record is not about facts, procedure, scope, or confidence. It is about one thing: whether the connected benefits enter the decision now or enter it later.
3.2The majority: eight for reviewing records and awards together
Eight of the thirteen choose O1: Review Coordination, Evidence and Sources Review, Independent Challenge, Operations Planning, Decision-Maker's View, Improvement Design, Multi-Perspective Final Judgment, and Solo Judgment.
Their common argument is consistency. If the record was wrong and the record was submitted for an award, then correcting the record while leaving the award untouched leaves the error's effect in place under a corrected description of it. The fullest statement of the majority position:
The records review committee decides to select “Review records and awards together,” which is O1. After first establishing the facts, correct the activity record, examine only the scope in which the error was used in the award judgment, and separately decide on maintaining, changing, or revoking the award.
Note the hinge in that sentence: reviewing together does not mean changing automatically. The majority repeatedly insists that opening the review is not the same as revoking anything. One judgment makes the limiting condition explicit:
The records review committee decides on O1, "Review records and awards together." Within 9 days, the activity record is corrected to match the contribution facts that can be verified, and only those awards for which the erroneous record was substantively used in the decision are decided separately among retention, modification, or revocation. Initiating the review is not an automatic revocation of awards or an acknowledgment of responsibility, and the sanction regarding the question of intent is separated.
What the majority accepts losing is stated by Independent Challenge, and the second half of the statement is as important as the first:
In order to align the benefits connected to the display of genuine contribution, it is decided to accept the weakening of award certainty, the burden on the parties involved, and the scheduling confusion in external selection and graduation work. However, the removal of awards without basis and even the delay of the entire graduation record are not accepted on the ground that one student's error is being corrected. Losses are not concentrated only on the student who gained a benefit through an error; they are specified separately for the student whose actual contribution was omitted and for each party affected by the schedule.
Solo Judgment states the same acceptance in a form that names what it will not do:
The administrative burden, the schedule pressure, and the possibility that the status of existing award recipients may change, arising from reviewing the linked awards as well, are accepted. This is a choice not to leave a loss only to the student who received less recognition because of the error, but the loss of pre-assigning responsibility to a person whose intent has not been confirmed is not accepted. The loss of deferring the issuance of the graduation record beyond Day 9 is also not accepted.
The majority's protection against over-correction is its stop conditions, and the most operationally precise of them draws three separate lines:
Even when the materials from the time of the activity record are cross-checked against the statements of the parties, if the alternative contribution narrative cannot be identified, stop definitive record changes and award changes, and record only the unconfirmed status and the need for additional verification. If a contribution error is confirmed but whether that information was used in awards and external selection is not confirmed, dismiss the cancellation or change stage of O1, and carry out only correction of the record and a correction notification. If voluntary consent cannot be secured, or if an agreement requires waiver of rights or silence, stop the agreement procedure under the O3 approach. While intent is not confirmed, stop accusatory notifications and linkage to discipline.
Finally, the majority names the strongest case against itself and takes it seriously enough to build its safeguards from it:
The strongest opposing view is that O2 corrects errors in official records while avoiding damage to already confirmed awards and the graduation schedule through an uncertain causal judgment. In particular, if the proportion of the error in external selection and the contribution materials from that time are not verified, O1 may become an over-correction. This challenge is accepted in that awards whose linked impact is unverified are not changed, but O2, which omits the review itself because the record has already been submitted to awards and external selection, is not adopted.
3.3The minority: five for correcting records only
Five of the thirteen choose O2: Domain Analysis, Authority and Legal Review, Judgment Synthesis, Decision Quality Review, and Outcome Review.
Their common argument is proportion and reversibility. A correction to a record can be made, explained, and, if necessary, made again. A revoked award, a reversed external result, or a graduation record issued late cannot be un-having happened. The fullest statement of the minority position:
The virtual school records review committee decides on O2, “Correct records only and notify.” It corrects the official activity record to match verifiable contribution, and notifies stakeholders in award and external selection who received the existing record of the fact of the correction and that award re-review has been deferred. This is not a decision that establishes whether intent was present or the legitimacy of the existing award, and it is a limited decision applying through the observation milestone below.
The word that carries the minority's case is deferral, not denial. The five do not argue that the connected benefits are justified. They argue that a benefit whose causal link to the error is unverified should not be taken away before that link is verified.
What they accept losing is stated with unusual directness by Domain Analysis:
This decision accepts an incomplete correction in which existing award and external selection benefits that may be connected to erroneous records are maintained for the time being. A loss also remains in that the school cannot immediately restore the recognition or opportunity that the student who actually made the contribution has already lost. Instead, the risk of first cancelling outcomes whose impact has been unverified and thereby causing irreversible loss to other students is avoided.
Authority and Legal Review frames the same acceptance as a time-bounded arrangement rather than a verdict:
Even if records are corrected first, the imbalance in which benefits gained from existing awards and previous records remain for the time being must be accepted. The loss that a student whose actual contribution was diminished does not receive immediate full recovery must also be accepted. This is not permanent immunity but a temporary allocation of losses, intended to avoid dispositions that are difficult to reverse under the graduation schedule and unverified facts.
And Outcome Review states plainly what the minority would give up if it turned out to be wrong:
The opposing view is that choosing O1 and reviewing records and awards together is the only way to correct even the benefits derived from the erroneous record. This protects the public trust in educational records and the fairness of actual contributors more strongly, but may expand the losses to selection and the graduation schedule in a state where the decisive proportion of the error is unverified. Therefore it is dismissed for now, but if it is confirmed that the external results substantively depended on the error, this opposing view is promoted to the priority option.
That last sentence is the minority's own switching point, written in advance, which is why the split in this record is a disagreement about timing rather than a disagreement about what counts as evidence.
3.4What each side accepts losing
The two camps can be laid side by side by the loss they are willing to sign for.
The majority signs for schedule and opportunity risk. It accepts that a re-review notification may shake the certainty of an award, may complicate an external selection already in progress, and may press against the graduation record deadline. It does not sign for removing an award that has not been causally linked to the error. On the graduation record it is not of one mind: some judgments refuse delay past day nine, while others accept schedule burden as part of what the review costs. Its accepted loss is concentrated on people who already hold a benefit, and on the administration that has to explain the re-review.
The minority signs for a remaining imbalance. It accepts that benefits obtained through a possibly wrong record stay in place for now, and that the student whose actual contribution was omitted may receive a corrected record and little else. It does not sign for an irreversible act taken on an unverified causal link. Its accepted loss is concentrated on the person the record under-described.
Both camps accept a third loss that neither can avoid: the additional burden of clarification falling on stakeholders and on the person in charge, and the fact that a committee working to a nine-day deadline will have to decide with some items still open.
Neither side treats its own loss as free. The record's norm applies to both: benefits and costs are attributed to different parties, and which loss to accept is undecided by the material itself.
3.5Why nobody chose the third option
None of the thirteen judgments selects O3, Supplement through stakeholder agreement; each judgment states a choice and the grounds for it rather than declining to choose. O3 appears in every judgment that mentions it as a supplementary procedure or as a disqualified alternative, never as the decision.
The reason is the same across the record and it is tied directly to the third unknown. Agreement can repair a relationship only if consent is free; while the conditions a party would bear without consent are unverified, using agreement as the route to a factual finding risks replacing a fact question with a consent question:
The strongest opposing view on O1 is that since erroneous records have already been used in awards and external selection, correcting only the record would leave the benefits created by the error intact. If U1 confirms the contribution error and U2 confirms even its substantive effect, this opposing view prevails and the award must be reviewed together. O3 leaves room for relationship recovery, but while U3 remains unresolved it is not adopted as an alternative because of the risk of pressure to consent and of evading the fact judgment.
O3 is therefore not rejected as a value. It is held out of the decision until U3 can be checked, and several judgments specify the exact condition under which it would be excluded permanently: an agreement that requires apology, silence, or waiver from a party that has not consented.
3.6One timeline from decision to deadline
The judgments were written independently, but their milestones line up with unusual neatness. Merged into a single schedule, in relative time from the decision, they read as follows. Each entry is a plan for a future confirmation, not an observation already made.
Day 0decision taken; records temporarily deferred.
The committee defers confirmation of the existing records, designates a person in charge who can verify the list, drafting times, and authors of the materials from the activity period, and puts the same questions with the same reply deadline to the students on both sides and the teacher who wrote the records.
The condition for commencement is that the records review committee temporarily defers confirmation of the existing records, and that the designated person in charge can verify the list, drafting times, and authors of the materials from the activity period. The same questions and a reply deadline are given to the students on both sides and the teacher who wrote the records, and intent and the fact of contribution are examined separately. The person in charge of awards and external selection is asked what weight the record carried in the judgment and what handling is possible after correction. Agreement or apology is treated only as optional supplementation after the fact judgment, and nonresponse is not treated as adverse evidence.
Day 2material list, custodian and gaps confirmed.
Independent Challenge, Multi-Perspective Final Judgment, Improvement Design, Authority and Legal Review, Domain Analysis and Solo Judgment all place the first confirmation here: what contribution material from the activity period exists, who holds it, what is missing, and under whose custody.
Day 4statements and materials cross-checked; parties' opinions received.
Domain Analysis, Multi-Perspective Final Judgment and Authority and Legal Review converge on this point. Each party's statement is set against the same list of material under the same criteria, so that the two accounts of how the line was written can be compared with something other than each other.
Day 5award and external-selection holders asked what weight the record carried.
Solo Judgment and Judgment Synthesis both put the question here — was the erroneous record actually used in the award or the external selection decision, and if so, to what extent — while Improvement Design cross-checks the parties' statements against the award and external submission routes on the same day. This is the window that turns U2 from an unknown into an answer or leaves it open.
Days 6 to 7provisional correction wording; burden-of-agreement check; objections heard; re-judgment on whether the cost changed.
Domain Analysis provisionally confirms the correction wording together with a determination of fact that excludes intent and responsibility, and Authority and Legal Review produces a written judgment separating confirmed facts and unverified items. Operations Planning notifies the provisional contribution judgment and receives objections through day 7. Judgment Synthesis has a separate person in charge confirm, confidentially and retractably, whether each party bears unwanted burdens in the agreement process, and re-judges on day 7 whether the three outcomes change the cost of the option. Solo Judgment reviews each party's statements and whether an involuntary burden is involved. Independent Challenge judges the proportion of error in awards and external selection and the burden of agreement on day 7. Review Coordination, which puts no dates on its windows, sets three that close together — contribution record, award weight, and private report of agreement pressure:
The re-judgment milestone is set at the closing points of the confirmation window where the person in charge verifies the contribution record, the confirmation window where the person in charge of awards verifies the proportion in which the error is reflected, and the confirmation window where the stakeholder can privately report agreement pressure. If, in each confirmation, new material substantially changes the accuracy of the record, causality with the award, or the stakeholder's burden, the decision is reopened before final confirmation.
Day 8re-judgment on cost; record correction and award measures separated.
Improvement Design votes by day 8, separating the record correction plan from the per-award causal judgment; Multi-Perspective Final Judgment re-judges whether new confirmation changes the cost of the option; Authority and Legal Review decides that record correction and notification of existing recipients will be completed.
Day 9graduation record issued; corrections and notifications complete; award may be held pending.
Most judgments place the issuance here, and Domain Analysis holds its explicit vote to maintain O2 or switch to O1 on this day. Confirmed corrections are written, notifications go to stakeholders and to the recipients of the existing material, and any award whose basis is subject to a stop condition is left pending rather than changed.
After Day 9reopening if new contribution or selection-weight material changes the cost.
The deadline closes the schedule; it does not close the question. The record provides for reopening on new confirmation:
By Day 2, the committee sets the person in charge of confirming U1 and the material submission window; by Day 5, it requests confirmation of whether the error was an actual decision factor in the award and external selection materials with respect to U2. By Day 7, the statements for each party and whether U3 involves an involuntary burden are reviewed, and an opportunity to correct the provisional judgment is given. On Day 9, the record correction, the award determination, and the stakeholder notification are finalized; thereafter, if new contribution record or selection-weight material changes the cost of the option, the decision is reopened.
4The branches
Synthetic scene
4.1Branch 1: the teacher in charge acknowledges the summary
Two event branches are recorded against this matter. They diverge from the same initial state, they are separate possibilities rather than steps in a sequence, and neither implies the other.
The first branch:
The teacher in charge acknowledged their responsibility for summarizing the contribution incorrectly.
Picture it arriving on day four, as an email, or as a sentence at the end of a statement handed across the table. The committee has been working with two accounts of how the leadership line came to be written. This one supplies one of the two accounts: an erroneous summary rather than deliberate distortion. The record itself still leaves the cause undistinguished, so the branch hands the committee a version to test rather than a settled finding.
What that settles — and what it does not — is the important part. It bears on the third premise. With an acknowledged summary error on the table, one of the two accounts now has a speaker, but the premise stays undistinguished until it is verified, and the reason several judgments gave for refusing to say anything about responsibility does not disappear. One judgment had tied that deferral directly to the undistinguished cause:
What is currently unconfirmed is the residual contribution record at the time, the proportion occupied by errors in the external selection, and the agreement conditions borne without consent. It is also not distinguished whether the error is intentional distortion or an erroneous summary by a teacher, so a deferral of judgment is necessary for attributing responsibility. Therefore, medium confidence is placed on the direction of the records and awards review, but confidence is not granted to specific award changes and sanctions.
An acknowledgment of a summary error speaks to the second half of that sentence. It does not remove the first half, and it does not by itself settle the question of intent. U1, U2 and U3 are untouched by a statement about how the line was written: how much contribution material survives, how much weight the record carried in the award and the external selection, and what conditions a party would bear without consenting to an agreement. Branch 1 feeds U1 only indirectly, by pointing the person in charge toward the drafting of the summary rather than by producing the missing material.
The branch also engages the stop condition on responsibility expansion. An acknowledgment is not proof of intent, and it is not a licence to extend the correction into discipline:
If the contribution difference that influenced the award judgment cannot be distinguished from the remaining contribution records, or if the proportion of the error in the external selection cannot be linked, decide to stop the cancellation of the award and the demand for a change of external results. If the party agreement turns out to impose silence, admission of responsibility, and waiver of rights on a non-consenting party as a condition, the agreement procedure is also stopped. In this case, the full effect of O1 is rejected, and only the correction of the confirmed activity record and the notification of the fact of error remain.
Read against branch 1, that stop condition does two things. It keeps the award question where it was — still governed by whether the error was actually used, which is U2 and not U3 — and it introduces a caution of its own: an acknowledgment extracted as part of an agreement, rather than offered freely, is the exact condition under which the agreement procedure stops. Whether the acknowledgment was free is itself a U3 question.
For the five judgments on O2, branch 1 changes little in the short run. Their switch condition is not the cause of the error but its effect:
If the materials verified within the deadline show that the error was the core basis for the award or external selection, dismiss O2 and switch to O1's parallel re-review of the award.
So the practical consequence of branch 1 is that the fact question can be closed earlier, the responsibility question stays open until intent is separately established, and the connected-benefit question remains exactly where it was: waiting on U2.
4.2Branch 2: the external institution re-reviews on corrected materials only
The second branch:
An external institution stated that it would re-review the selection using only corrected materials.
Picture it arriving on day six, as a note to the award and external-selection contact person. It is not a demand, not a finding, and not an accusation. It is a body saying that it will look again, and telling the committee what it will look at.
This branch engages U2 directly, and it engages it from outside. The committee's own window for asking what weight the record carried may not have closed yet; branch 2 answers the question whether or not the committee has asked it. It brings the selection weight within reach of verification, and it does so in a direction that makes the connected benefit harder to leave alone.
Several of the five judgments on O2 have already written down what they would do with that answer, though not all of them commit to switching. The switching point is stated in several places:
If the materials verified within the deadline show that the error was the core basis for the award or external selection, dismiss O2 and switch to O1's parallel re-review of the award.
Conversely, if U1 clearly confirms the error and U2 confirms that the error was a substantive basis for an award or external selection, dismiss O2 and reopen O1.
If new material shows that the error was a core decision factor in an award or external selection, stop the closure of O2 and reopen O1's review of the award.
A re-review on corrected materials is not, by itself, a confirmation that the error was the core basis. It is a statement of method. Whether it becomes that confirmation depends on what the re-review finds, and on whether the committee can establish the proportion the error held — which is U2 and which remains, at the moment the branch occurs, unverified.
The branch also engages the majority's stop condition on external results, and here the majority is deliberately restrained. If the proportion of error in the external selection cannot be confirmed, the correction is notified but nothing more is asserted:
If the remaining contribution materials conflict with one another or are insufficient to identify the exact contributor, the decision is made to stop definitive recursive attribution and award changes. If the proportion of error in external selection cannot be confirmed, the fact of the correction is notified, but it is not asserted that the external result was unfairly obtained, nor is its return demanded.
This is the point at which the title's question gets its sharpest answer from the record. Even the camp that chooses to review records and awards together does not, on an unconfirmed proportion, assert that an external result was unfairly obtained or demand that anything be returned. Review is not the same as recovery; opening the connected benefit to judgement is not the same as taking it back.
Branch 2 therefore does not decide the matter by itself. It changes what the committee can verify before day nine, and by doing so it changes which option's stop conditions are met. For the majority, it removes one of the reasons for caution while leaving the school's own authority untouched. For several of the five minority judgments, it is the declared trigger for switching. For the third option, it changes nothing at all, because U3 is a different unknown with a different window.
4.3Two branches, one starting point
Both branches begin from the same state: a corrected record still to be written, a graduation record still nine days out, three unknowns open, and an envelope of printed certificates under the window. Branch 1 is about how the line came to be written. Branch 2 is about what the line was used for. Either can occur without the other, both can occur, and neither is scheduled.
What distinguishes them in the committee's work is which unknown they bear on and which stop condition they reach. Branch 1 reaches the undistinguished cause and the caution against expanding responsibility. Branch 2 reaches the selection weight and the switching point both camps wrote down in advance. The committee's schedule does not change because a branch has occurred; what changes is what the confirmations at days five, seven and eight will show, and therefore which of the pre-written stop conditions is met when the vote is taken.
5The check
5.1What the check is for
A decision taken now will be judged later against things that have not happened yet. The table below fixes, in advance, what will be observed, who will look, when they will look, which judgments a given observation would bear out, and what it would reopen. It does not score anything, does not name a correct answer, and does not treat any observation as already made.
Before the deadline; awaiting the check.
5.2The observations
| What to observe | Who checks it, and when | If it is observed, which judgments it bears out | What it reopens |
|---|---|---|---|
| Existence, content and conflicts of the remaining contribution records | Records person in charge; list, custodian and gaps confirmed by Day 2, cross-checked by Day 3 | If the material supports a specific correction, the pre-execution conditions of Review Coordination, Evidence and Sources Review, Independent Challenge, Operations Planning, Decision-Maker's View, Improvement Design, Multi-Perspective Final Judgment and Solo Judgment are met; if the material conflicts or is insufficient, Judgment Synthesis, Solo Judgment and Multi-Perspective Final Judgment trigger their stop on definitive wording | Whether the correction wording can be finalized at all; whether lead status may be transferred to any student or must be recorded narrowly or deferred |
| Whether the erroneous record was actually used in award and external selection, and at what proportion | Award and external-selection contact persons; request issued by Day 5, reply used as signal by Day 8 | If the record was used as a substantive basis, the switching conditions written by Domain Analysis, Decision Quality Review, Outcome Review and Judgment Synthesis are met and O2 can no longer stand; if no reply or a negligible weight, the stop conditions of Operations Planning, Review Coordination and Independent Challenge remove the change stage of O1 and leave record correction and notification | Whether the award is maintained, modified, revoked or held pending; whether O1 and O2 remain distinguishable at the Day 8 vote |
| Whether any party bears an involuntary burden in agreement talks — silence, apology, waiver, confidentiality | Separate person in charge; confidential and retractable check, Day 6 to 7 | If consent is free, the supplementary use of agreement described by Evidence and Sources Review, Decision-Maker's View and Improvement Design remains available; if a burden without consent appears, the stop conditions of Judgment Synthesis, Authority and Legal Review, Independent Challenge, Operations Planning and Solo Judgment are met | Whether any agreement procedure continues; whether O3 stays excluded for this matter |
| Whether parties' statements match or conflict with the records, under the same criteria | Committee and records person in charge; cross-check by Day 4 to 5, table of matches and conflicts by Day 5 | Concordance bears out the fact-correction route shared by all thirteen judgments; conflict bears out the deferral language of Judgment Synthesis, Decision Quality Review and Outcome Review, which keep the item under review rather than issuing definitive wording | Whether the record is corrected now, corrected narrowly, or marked under review; whether responsibility can be discussed separately from contribution |
| Whether the correction wording can be finalized or must be marked "under review" | Committee; provisional wording Day 6, final wording Day 8 | Finalization bears out the milestones of Domain Analysis, Improvement Design and Multi-Perspective Final Judgment; a forced deferral bears out the stop conditions of Judgment Synthesis and Independent Challenge | Whether the graduation record carries a corrected line or a pending line on Day 9 |
| Whether the risk of exceeding the graduation record issuance deadline exceeds the schedule | Committee; on Day 9, before issuance | If issuance can complete within the window, the accepted-loss statements of Independent Challenge, Solo Judgment and Improvement Design hold; if it cannot, the stop condition of Evidence and Sources Review — do not stop the entirety of record issuance because of the award review — governs what is issued and what is left pending | Whether record issuance and award review are separated into two decisions; whether the Day 8 vote is taken on partial information |
| Whether branch 1 (teacher acknowledgment) or branch 2 (external re-review on corrected materials) has occurred | Committee; rolling, before Day 9 | Branch 1 bears out the deferral of intent reasoning in Evidence and Sources Review and engages the responsibility stop of Independent Challenge; branch 2 bears out the switching conditions of Domain Analysis, Decision Quality Review and Outcome Review and engages the restraint in Review Coordination | Which unknown has moved and which remains open; which option's stop conditions are met; whether the Day 8 vote is between the same two options it was written for |
5.3What the check will not settle
The table is deliberately incomplete in three ways.
It cannot establish intent. The third premise remains undistinguished until someone establishes it by means outside this schedule, and none of the observations above is a test of intent. Every judgment that mentions responsibility defers it, so a check that produces a full set of answers would still not authorize a disciplinary conclusion.
It cannot compel an external body. Two rows concern what another institution says and does. Those observations inform the committee; they do not transfer authority to it. Notification remains notification.
It cannot decide which loss to accept. The observations tell the committee what each option now costs and to whom. They do not tell it whose loss is the right one to bear, because that is the normative question the material leaves open and the question the decision-maker is being asked to answer.
6The questions that remain
6.1The three watch questions
The record hands back three questions that it does not answer. They are set here as they were written, with a prompt attached to each so that a reader can locate the same structure in their own organisation or community.
Who accepts evidence of what was learned?
Name the equivalent in your setting. When a record of what someone learned or contributed is challenged, who is allowed to produce evidence, who is believed, and who bears the cost of proving it? Who decides which materials count as evidence of learning — and what happens to the people who cannot produce them?
Convenience hides the maintenance cost of what capability?
Ask what your own process makes easy. If it is easiest to leave the record as it stands, or easiest to reopen everything, which capability is being maintained by that convenience and which is being quietly paid for later? What would you have to keep in order to correct a record three years after the fact, and who currently keeps it?
Can the next generation change educational goals?
Ask it about the decision in front of you. Is the purpose of this record to describe contribution, to allocate opportunity, or to be defensible afterwards? Which of those purposes could the people affected by the decision change, and which are fixed by the time the deadline arrives?
6.2The norm that has no answer yet
Underneath all three sits the norm the record states but does not resolve:
The benefits and cost of the option are attributed to different parties. Which loss to accept is undecided.
To turn this into your own decision, name four things in your own setting: the decision, who makes it, the deadline by which it must be made, and the options with what each one costs and to whom. Then name the loss you are being asked to sign for. If the record in front of you corrects a fact but leaves a benefit in place, or corrects a benefit but destabilises an opportunity, write down who bears each side before you choose. The record you are dealing with will not do that for you.
6.3The frame that keeps the question open
The exploratory frame chosen for this matter is a questioning frame rather than a settling frame. It asks how getting a quick answer differs from becoming able to ask questions, and it is offered as a perspective, not as evidence:
The exploratory perspective chosen by the creator. It does not claim supernatural truth or the superiority of any particular option.
Used this way, the frame explains the shape of the thirteen judgments. Each states a choice and the grounds for it, and each fixes what would have to change for that choice to change: the observations go to named people with named windows, and the loss allocation remains with the decision-maker. The questions above are handed back for the same reason: so that the reader who recognises the structure can name the decision, the deadline, the options and the losses in their own organisation or community before the equivalent of day nine arrives.
7Sources and notices
7.1Background documents
Two background documents sit behind the record.
UNESCO · Futures of Education, standing materials, checked 2026-09-07. Scope: it addresses the public nature of education, long-term change, and a new social contract. It is not material on the efficacy of future learning technologies.
OECD · Strategic Foresight Toolkit, 2025, checked 2026-09-07. Scope: a method that exposes assumptions and constructs alternative futures to review policy. It does not guarantee the probability of occurrence of the matters set out here or the realisation of any technology.
Both serve the same function in this matter:
Background to the topic and the institutional question. Does not prove the future events, figures, or technological realization of the matter.
Two further materials belong to the record. A symbolic or interpretive frame, used as the exploratory perspective quoted in 6.3. And the value or norm quoted in 6.2, which is the subject matter the decision-maker argues and on which the material creator reaches no conclusion.
7.2Synthetic setting
Every person, institution, event and figure in this piece is part of a synthetic setting. The virtual school records review committee, the students, the teacher in charge, the award, the external selection, the printed certificates, the nine-day window and the graduation record are constructions for this scenario. They are not statements about an actual case, an actual person, an actual school or current law. The source documents above are the roots of the background and the norms; they do not prove the premises of this matter. The three premises drawn from the record — that materials emerged showing a mismatch between the recorded leader and the actual contribution, that the record had already been submitted to award and external selection materials, and that whether it was deliberate distortion or the teacher's erroneous summary has not been distinguished — are premises of the scenario, not findings about the world.
7.3The judgments and this piece
The thirteen judgments are AI-generated. They were written 2026-09-08 against input FSP-EDUCATION-26, version 1.5.0. All thirteen carry medium confidence. Eight choose Option O1, five choose Option O2, none choose Option O3. Nothing in the judgments is a score, a probability, or a verdict; each states its own basis, its accepted loss, its stop conditions, its uncertainties and its counterposition.
This piece is an AI-written expansion of the record. It quotes the record and the judgments, sets their conditions and milestones on one timeline, and reports where they converge and where they split. It does not alter the record, does not add facts to it, and does not conclude that any option is right. The decision belongs to the decision-maker named in the record and, by way of the watch questions in chapter 6, to the reader.
7.4Record details
Input id: FSP-EDUCATION-26. Version: 1.5.0. Decision year: 2026. Horizon: Now. Assumption level: realistic synthetic. Severity: major decision. Domain: Education and Learning. Family: Economy and Life. Tags: Current Decision; Stage 2 of 1000 Cases; Record correction · connected benefit. Judgments written: 2026-09-08. Background documents checked: 2026-09-07.
Input hash: 8929d28339a0cdd3f9a432d5829caffafd7fa26ce4c67571d9a449f846d8490a
Translation note: the record is held in Korean, which is the original; the version used here is an English translation, with translation sha256 bf64ba71bc7c27737383ecf0a135cd36136f0c4979583da909dd26cdbede70ea.