False-Positive Harm Budget Dashboard¶
Monitoring dashboard — instantiates Self-Targeting Defense Guardrail
Meters the running cost of wrongful self-engagements against a pre-set allowance, weighting each by harm intensity, so the defense's autoimmune damage is priced and capped.
Enforcement systems are usually measured by threats caught and almost never by legitimate parties wrongly harmed — which is precisely how an autoimmune failure hides in plain sight. False-Positive Harm Budget Dashboard fixes the scoreboard: it assigns a harm budget — an explicit allowance of wrongful self-engagement the system may spend over a window — and meters live consumption against it, so wrongful harm becomes a first-class, visible, capped quantity rather than an untracked side effect. Its defining move is that each confirmed false positive is not counted as one tick but priced by its harm intensity and reversibility, so an irreversible wrongful destruction draws down far more budget than a reversible flag, and the running total is compared to the cap in real time. It is a continuous production accounting instrument for a defense that is already live; it does not evaluate a defense before deployment and it does not itself characterize the classifier's error profile — it prices confirmed harm and watches the meter.
Example¶
An e-commerce marketplace runs an automated seller-suspension system, and wrongful suspensions of legitimate sellers are its autoimmune cost. The False-Positive Harm Budget Dashboard sets a weekly allowance and meters against it. Each confirmed wrongful suspension is priced by harm: a two-hour reversible hold on a hobby seller draws a little; a week-long suspension that made an established seller miss a peak sales event, hard to undo, draws heavily. As overturned appeals confirm false positives through the week, the meter climbs. When the priced total approaches the cap, the dashboard signals that the suspension system is spending its self-harm allowance too fast — prompting a tightening of thresholds, and, if the cap is breached, tripping an automatic halt of automated suspensions pending review.
Setup to outcome: instead of "we caught lots of fraud," leadership sees "we have spent 80% of this week's wrongful-harm budget by Wednesday, mostly on a few high-severity cases." The distinction that matters is that the harm was priced and capped, not merely counted.
How it works¶
- Set an explicit allowance. A harm budget is fixed for a window — how much wrongful self-engagement the defense may spend — turning an implicit tolerance into a stated, governable number.
- Price each false positive by harm. Confirmed wrongful engagements are weighted by intensity and reversibility, so severity and permanence determine budget draw, not a flat per-incident count.
- Meter live against the cap. The running priced total is displayed against the allowance in real time, making the approach to the ceiling visible before it is crossed.
- Escalate on approach and breach. Nearing the cap prompts tightening; breaching it triggers a defined response — throttle, require higher authorization, or halt — so the budget is a control, not just a chart.
Tuning parameters¶
- Budget size — how much wrongful harm is permitted per window. A generous budget keeps the defense aggressive against threats but tolerates more autoimmune damage; a tight budget protects self but throttles enforcement sooner.
- Harm pricing weights — how steeply intensity and reversibility scale each draw. Steep weights make irreversible harm dominate the meter (protective of the worst cases) but can let many minor false positives pass unremarked.
- Window length — the accounting period over which the budget resets. Short windows react fast to spikes but are noisy; long windows smooth noise but let sustained bleed accumulate.
- Breach action — what crossing the cap does, from alert-only to hard halt. A hard halt guarantees the ceiling holds but hands adversaries a lever to disarm the defense by manufacturing false positives.
When it helps, and when it misleads¶
Its strength is that it prices the archetype's most under-measured cost and puts a ceiling on it — the direct antidote to "enforcement success measured without pricing wrongful self-engagement." Borrowing the error-budget[n1] discipline, it converts a vague aspiration ("don't harm legitimate users too much") into a spent-or-remaining number that forces an explicit trade between coverage and self-harm, visible to everyone before an incident becomes a crisis.
Its failure mode is that a budget is only as honest as its inputs. It can only price confirmed false positives, so wrongful harms that are never detected or appealed never hit the meter, and the dashboard reports a comforting low number while real autoimmune damage runs off the books. Miscalibrated harm weights distort the picture — undervalue irreversibility and the worst cases barely register. Two classic misuses: gaming the metric by discouraging appeals so fewer false positives get confirmed, and letting a breach-halt become an adversary's disarm button. The guarding discipline is to feed the meter from independently-confirmed harm (upheld appeals, audits) rather than self-reported counts, and to revisit the harm weights so the budget tracks true damage, not the damage that happens to be easy to see.
How it implements the components¶
harm_budget_for_false_positives— it is the budget: an explicit allowance of wrongful self-engagement, metered live and enforced at the cap.harm_intensity_and_reversibility_map— it prices each confirmed false positive by intensity and reversibility, so budget draw reflects severity rather than a flat count.
It does not evaluate a defense before it is live or characterize the classifier's error profile — running an enforcer in observe-only to measure its would-be threat_classifier_with_uncertainty_state behavior and probing its adversarial_spoofing_and_evasion_model robustness belong to Shadow Mode and Canary Enforcement; shadow-and-canary vets a defense before rollout, this dashboard meters one already in production.
Related¶
- Instantiates: Self-Targeting Defense Guardrail — supplies the priced, capped accounting of wrongful self-engagement the archetype requires in production.
- Sibling mechanisms: Appeal and Rapid Restoration Workflow · Engagement Kill Switch · Graduated Response Matrix · Post-Incident Autoimmune Review · Protected-Self Allowlist with Expiry · Quarantine-Before-Destroy Rule · Self-Status Cross-Check · Shadow Mode and Canary Enforcement · Two-Key High-Harm Engagement
Editorial Notes¶
Form Classification¶
Form family: Monitoring, Sensing & Alerting
Rationale: False-Positive Harm Budget Dashboard operates as an ongoing sensing arrangement that repeatedly observes actual state and surfaces changes or alerts because it meters the running cost of wrongful self-engagements against a pre-set allowance, weighting each by harm intensity, so the defense's autoimmune damage is priced and capped.
Independent corroboration: The frozen evidence defines False-Positive Harm Budget Dashboard as 'Meters the running cost of wrongful self-engagements against a pre-set allowance, weighting each by harm intensity, so the defense's autoimmune damage is priced and capped', so its operative form is Monitoring, Sensing & Alerting.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Military & Strategic Studies
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Specialized
Rationale: Military doctrine treats fratricide as harm caused by a defensive force's own target-classification and engagement process, requiring identification controls and explicit risk mitigation.
Related originating lineages:
- Economics & Finance — Budgeting a cumulative risk allowance materially shapes the running cost ceiling.
- Engineering & Design — Safety engineering materially contributes harm ceilings and monitored guardrails.
- Security Studies & Intelligence Analysis — Measuring wrongful engagements and self-targeting damage is rooted in defense, security, and threat-detection operations.
- Statistics & Experimental Design — False-positive measurement materially supplies the running error account.
- Ethics of Technology & AI Governance — Human-impact accounting for false positives materially shapes harm weighting and governance caps.
Review resolution: Army protection doctrine documents the trade between fratricide prevention and operational initiative. Reliability error budgets and statistical false-positive measurement are later components of the synthesized dashboard.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
Notes¶
The dashboard is an accountant, not a decision-maker: it prices and caps wrongful harm but does not itself change thresholds or halt anything except through the breach action it is wired to. Its meter is most useful as the incident stream feeding Post-Incident Autoimmune Review and as one of the automatic trip conditions for Engagement Kill Switch — the number is the input; the architectural change and the stop live elsewhere.
[n1] An error budget is an explicitly allowed amount of failure over a period — for example, the small fraction of downtime permitted under a reliability target. Spending it forces an honest trade between moving fast and staying reliable. This dashboard applies the same accounting to wrongful self-engagement: a stated allowance that, once metered and priced, disciplines how aggressively the defense may act. ↩