Compassion Benchmark

Viewing archive: May 13

Compassion BenchmarkWednesday, May 13, 2026No. 29

Daily Briefing

Hungary crosses from Developing to Functional (37.5 → 41.4) — first material upward band crossing in the countries index this year.

Entities monitored
1,160
Fully assessed
11
Score changes
0
Risk signals
1
Top score changemedium

Hungary: 37.5 → 41.4 (Δ +3.9)

What happened

Hungary MAGYAR-ERA FIRST OPERATIONAL BASELINE — BAND CROSSING: Developing to Functional (37.5 → 41.4). Cabinet decentralized to 16 ministries with 3 new standalone portfolios (health, environment, education); comprehensive audit of ministries and state-owned companies ordered; payment suspension beyond normal operational management enacted; first cabinet meeting held May 13 with drought-response expert consultation and emergency measures.

Why it matters

ICC arrest-warrant enforcement reversal credited on INT dimension. Conservative-rounding discount applied from full math reconstruction (43.0) to boundary-case-respecting 41.4.

Countriesfunctionalmedium confidence

Today's analysis

The most significant editorial findings in the May 13 briefing.

01

Hungary's Developing to Functional crossing (37.5 to 41.4) is the first material upward band crossing in the countries index in 2026. This is an institutional milestone: a 16-year authoritarian government replaced by a pro-democracy one, with enacted Day-1 actions (cabinet decentralization, audit order, ICC reversal) sufficient to cross the Functional threshold under conservative scoring. The band crossing is real. The 0.4-point buffer above the floor is narrow. 30-day re-assessment June 9.

02

xAI/Grok floor confirmed with the largest concurrent multi-jurisdictional regulatory action stack in AI Labs index history: 8+ jurisdictions, concurrent formal proceedings, anti-regulatory litigation posture rather than compliance. RAND frames it as a 'regulatory reckoning,' not a routine enforcement cycle. This is qualitatively distinct from prior AI lab floor designations.

03

Russia's complete bad-faith ceasefire-format cycle is now the most extensively documented conduct pattern in the entire benchmark. Three phases (prepare, strike, sustain) spanning May 9-13 establish a formal template for how state actors can instrumentalize ceasefire formats to maximize harm capacity. This methodology is now available as a reference template for any future ceasefire assessment.

04

Myanmar's school airstrike death toll rose to 25 on Day 2 — one of the deadliest attacks on children since the 2021 coup. ~100 children were present during the strike. International aviation fuel embargo call is the most significant international accountability mechanism activated this cycle.

05

Sudan received its first coordinated AU+UN joint condemnation (Banjul Joint Declaration, May 12): a milestone that raises floor-exit evidentiary burden substantially, as future rehabilitation would require engagement with both AU and UN mechanisms simultaneously.

06

Israel's MSF manufactured-malnutrition documentation (90% premature births, 84% low birth weight from malnourished mothers) constitutes the highest-quality clinical evidence tier in the current floor record — independent clinical observation from MSF medical teams with direct patient-level data.

07

A new positive-conduct category — 'external-accountability-reversal' — is introduced for the first time in the framework's history. This is a methodology evolution moment: all prior new categories documented harmful conduct. The framework can now formally credit institutional actions that strengthen external accountability systems.

Signal stack

4 signals
medium

Countries — Pakistan Monsoon Emergency

NDMA issued pre-monsoon weather alert for major weather system affecting most of Pakistan May 13-17.

medium

Countries — Hungary Structural Change (UPWARD TRAJECTORY)

FIRST POSITIVE SECTOR ALERT IN MANY CYCLES.

medium

AI Labs — xAI Multi-Jurisdictional Regulatory Enforcement

xAI/Grok floor confirmed with 8+ jurisdictions in formal regulatory proceedings — largest concurrent regulatory action stack in the AI Labs index.

medium

Countries — Post-Ceasefire Conflict Zone Cluster

Four simultaneous floor entities with new conduct evidence this cycle: Russia (complete bad-faith ceasefire-format cycle documented, Day 2 post-format surge 100+ drones); Myanmar (school airstrike death toll rose to 25); Sudan (first coordinated AU+UN joint condemnation, 880+ drone deaths documented); Israel (MSF manufactured-malnutrition crisis, 2,000+ ceasefire violations milestone).

Score change detail

Full evidence record for entities with score changes in this cycle.

Hungary

Countriesmedium confidence
37.541.4
+3.9 pts
developingfunctional

Hungary MAGYAR-ERA FIRST OPERATIONAL BASELINE — BAND CROSSING: Developing to Functional (37.5 → 41.4). Cabinet decentralized to 16 ministries with 3 new standalone portfolios (health, environment, education); comprehensive audit of ministries and state-owned companies ordered; payment suspension beyond normal operational management enacted; first cabinet meeting held May 13 with drought-response expert consultation and emergency measures. ICC arrest-warrant enforcement reversal credited on INT dimension. Conservative-rounding discount applied from full math reconstruction (43.0) to boundary-case-respecting 41.4. 30-day re-assessment June 9.

Evidence record
  1. 16-minister cabinet sworn in; Sulyok formally appointed all 16 at Sándor Palace; cabinet decentralized to 16 ministries.
  2. May 13 first cabinet meeting; expert consultation on drought; immediate audit of ministries and state-owned companies; payment/commitment suspension; rule-of-law and Western-alliance priorities.
  3. Government formation pace and ministerial appointments corroboration.
  4. Magyar ICC enforcement policy reversal — direct contradiction of Orbán's non-enforcement during April 2025 Budapest visit. External-accountability-reversal conduct category first application.
  5. Third-party endorsement of Magyar ICC-enforcement posture sustained through April-May 2026.
  6. Independent analysis of Magyar government rule-of-law restoration agenda scope.
Boundary watch resolution

New band crossing: Developing to Functional. Proposed composite 41.4 sits 0.4 above Functional floor (41.0). Boundary-case protocol active. Conservative-rounding rationale: full mathematical reconstruction yields 43.0; assessor applied translation-from-promise-to-action discount, crediting only enacted institutional actions (not stated intent), pinning at 41.4. 30-day re-assessment scheduled June 9.

Next assessment triggers
  • May 25-26: Magyar-von der Leyen summit
  • May 27: EU funds plan submission deadline (ACT/SYS evidence)
  • May 31: Sulyok dismissal compliance deadline (ACC trigger)
  • June 9: 30-day re-assessment — Functional band consolidation decision point
  • June 12: Translation-from-promise-to-action check on boundary case (41.4 vs 41.0 floor)
  • Aug 31: EU rule-of-law milestone deadline (statute repeals due)
  • Watch: anti-LGBTQ statute repeal introduction (EQU credit pending)
  • Watch: audit findings publication (ACC anchor 3-4 credit possible)
  • Watch: ICC-warranted person entry test

Score movements

All entities assessed this cycle. No score changes.

11 assessed
Countries
37.541.4+3.8999999999999986
functionalmedium
Countries
0
critical
Boundary watch5 entities near a band threshold

Entities approaching band boundaries

Countries
17.5
POST-BAND-CROSSING CARRYFORWARD — band crossing (Developing to Critical) realized May 12. Watch cycle 1 completed May 13. Active monsoon weather watch through May 17.
Countries
41.4
RESOLVED AS UPGRADE — Developing to Functional. Prior composite 37.5. Boundary case at 41.4 vs 41.0 Functional floor (0.4 buffer). 30-day re-assessment June 9. Translation-from-promise-to-action watch active.
Robotics Labs
81.4
MATH-HYGIENE CARRYFORWARD — published 81.4 (Exemplary) vs reconstructed 73.4 (Established). +8.0 discrepancy largest in robotics-labs cluster. Band may shift to Established on math-hygiene resolution. Data-team escalation.
Countries
37.5
CARRIED FROM MAY 12 — post-band-crossing downward trajectory. HIV-service criminalization compound deepening. First criminal conviction (6-year sentence) enforcement milestone. Downward pressure active.
Countries
35.9
CARRIED FROM MAY 12 — internal reconstruction 0.9 below Functional floor. UNGA May 20 vote prospective positive event. +1.6 carry-forward dimensional credit active.

Evidence ledger

Primary sources reviewed in this briefing cycle. 6 sources linked.

Primary sources reviewed in this briefing: domain, source type, entity linked, dimension, and external link.
SourceTypeEntityDimensionLink
washingtonpost.comNewsHungarySYSOpen
english.news.cnSourceHungaryACTOpen
intellinews.comSourceHungarySYSOpen
aljazeera.comNewsHungaryINTOpen
hrw.orgNGOHungaryINTOpen
ipsnews.netSourceHungarySYSOpen

Risk signals

Developments that may affect future scores. Watch items from the May 13 briefing.

Risk

Pakistan monsoon-impact carryforward

Pakistan entered Critical band May 12 and immediately faces a multi-day elevated weather emergency (NDMA May 13-17 alert; 2026 monsoon forecast 22-26% above normal). If mass-casualty flooding materializes in the May 13-15 peak window, the EMP dimension could generate further downward pressure on a composite already in Critical. Pakistan at 17.5 sits 17.5 points above the floor — not at immediate floor risk — but the trajectory is downward and climate-vulnerability is an uncontrolled variable.

Window2026-05-13 to 2026-05-17
Risk

Apptronik math-hygiene resolution risk

Published 81.4 (Exemplary) vs reconstructed 73.4 (Established) represents a +8.0 discrepancy now at 7 cycles without resolution — largest in the robotics-labs cluster. When the formula audit resolves this, Apptronik will shift from Exemplary to Established band. This is not a compassion-behavior risk, but it is a publication-integrity risk: the published score currently overstates Apptronik's placement by one band category.

WindowPending formula audit resolution
Risk

Hungary translation-from-promise-to-action

Hungary's 41.4 Functional placement sits 0.4 above the Functional floor. The assessor applied a conservative discount from the full mathematical reconstruction (43.0), crediting enacted Day-1/Day-2 actions but holding stated commitments (statute repeals, EU fund unlock, Sulyok dismissal, anti-LGBTQ law repeal) as unconfirmed. If Sulyok dismissal compliance fails (May 31 deadline) or EU fund plan falls short, the June 9 re-assessment could pin Hungary back toward the boundary. Upside: if all May 27-31 milestones deliver, full 43.0 consolidation is plausible.

Window2026-05-27 to 2026-06-09
Risk

Open Bionics formula audit blocking publication integrity

Open Bionics now at 11 cycles without formula audit resolution. Published 97.5 carries a -10.0 reconstruction discrepancy that has been documented since cycle 1. At 11 cycles, this is no longer a maintenance item — it is an active publication-integrity risk. Every cycle this remains unresolved, the benchmark publishes a score known to be materially incorrect by a full 10 points.

WindowImmediate — audit must begin this week

Failure modes in this briefing

Recurring patterns the ACB methodology tracks as structural barriers to institutional compassion. Detected from evidence documented in this cycle.

Failure mode

Stated commitment operational hollowing

Public commitments are maintained in language while the operational machinery to fulfill them is dismantled, under-resourced, or conditionally applied. The commitment becomes a rhetorical position rather than a behavioral constraint.

Detected inHungary translation-from-promise-to-action
Methodology innovation1 new conduct category

New analytical categories

The ACB framework is extended when conduct patterns appear that existing categories cannot capture. Each new category is dated and tied to its first-application entity, creating an auditable record of framework evolution.

Draft

external-accountability-reversal

A positive-conduct category applied when an institution actively reverses a prior policy of non-compliance with or withdrawal from an established external accountability mechanism, thereby strengthening the mechanism's institutional integrity and expanding its operational reach. Requires: (a) a prior documented non-compliance or withdrawal posture, (b) an enacted policy reversal (not merely stated intent), and (c) practical effect on the accountability mechanism's reach.

First applied tohungary

Confirmed positions

Entities reassessed for this briefing where published scores remain supported by current evidence.

Confirmed positions from the May 13 briefing.
EntityIndexBandPublishedAssessedDeltaDateFinding
Countriescritical17.517.50Pakistan POST-BAND-CROSSING WATCH CYCLE 1 — confirmed at 17.5 Critical. No new mass-casualty events in May 13 window. NDMA pre-monsoon preparedness (PM-approved strategic plan May 7) noted as positive sub-threshold AWR/ACT offset. Active weather watch through May 17; EMP dimension downgrade possible if mass flooding materializes May 14-17.
Countriesdeveloping23.423.40Nigeria rotation-backfill confirmed at 23.4 Developing. 8-day gap closed. Math-hygiene resolution from May-05 baseline verified clean (dimension means reconstruct to 23.4). Boko Haram/JAS Borno resurgence, ISWAP operations, and security-force-abuse pattern sustained; VAPP Act review and Borno Police Commissioner anti-torture warning are institutional positive counterweights.
CambodiaFirst baseline
Countriescritical12.512.50Cambodia FIRST AGENT BASELINE confirmed at 12.5 Critical. Scanner's reported 7.5 was stale; canonical 12.5 confirmed by dimension-mean reconstruction (all 8 dimensions at 1.5). Dominant negative signals: 53 active scam compounds (Amnesty Jan 2026); 100K-150K trafficked workers across SE Asia; Tier 3 TIP placement; Hun Manet authoritarian succession; CNRP opposition dissolution legacy; press freedom collapse.
Countriescritical000Sudan FLOOR CONFIRMED at 0.0 Critical — floor conduct documentation updated with two major primary-source events: UN HRC drone-casualty data (drones caused 80%+ of civilian deaths, 880+ killed in first 4 months of 2026) and Banjul Joint Declaration (first coordinated AU+UN formal condemnation of both warring parties).
Countriescritical000Israel FLOOR CONFIRMED at 0.0 Critical — MSF 'manufactured malnutrition crisis' documented (50%+ pregnant women malnourished, 90% babies premature, 84% low birth weight); 2,000+ ceasefire violations milestone; UNRWA 14-month access block sustained; Hungary ICC reversal adds external INT-dimension accountability signal narrowing travel footprint.
Countriescritical000Russia FLOOR CONFIRMED at 0.0 Critical — Day 2 of post-format-offensive-surge: 100+ drones, 8+ killed in civilian-area barrage, railway infrastructure and civilian sites deliberately targeted. Kremlin confirmed 'special military operation is continuing.' Complete bad-faith ceasefire-format cycle now fully documented across May 9-13 (three phases: prepare + strike night 1 + sustain night 2).
Countriescritical000Myanmar FLOOR CONFIRMED at 0.0 Critical — Day 2 update on May 12 school airstrike: death toll rose to 25 (2 additional students aged 7-8 died from injuries); ~100 children at school during strike; 100+ wounded. One of deadliest attacks on children since 2021 coup. Fortify Rights and Malaysia call for ASEAN aviation fuel embargo on junta.
xAI/GrokFirst baseline
Ai Labscritical000xAI/Grok FLOOR CONFIRMED at 0.0 Critical — multi-jurisdictional regulatory enforcement cluster (8+ jurisdictions: UK ICO, UK Ofcom, EU DSA, California AG, 35-state coalition, Canada Privacy Commissioner, Brazil halt, xAI v. Colorado AI Act with DOJ intervention) documented as densest concurrent regulatory action stack in AI Labs index. RAND: 'Grok Isn't a Glitch — It Is a Regulatory Reckoning.' Anti-regulatory litigation posture (Colorado AI Act suit) rather than compliance. No safety-reform program in trajectory.
ApptronikCarry-forward +-8
Robotics Labsexemplary81.481.40Apptronik rotation-backfill CONFIRMED at 81.4 Exemplary. Math-hygiene +8.0 discrepancy (published 81.4 vs reconstructed 73.4) re-asserted as largest in robotics-labs cluster — data-team escalation. No new material score-moving evidence this cycle. Apollo commercial pilots (Mercedes-Benz, GXO, Jabil) continuing.
Robotics Labsestablished60.960.90Agility Robotics rotation-backfill CONFIRMED at 60.9 Established. Compound-positive baseline sustained: Toyota Canada Feb 2026 RaaS deployment, 2022 anti-weaponization pledge, ASTM safety-standards leadership. ASTM May 27-28 Boston convening upcoming. Math reconstruction clean (60.9 = 60.9). No boundary case.

Floor conduct record

Cycle-specific conduct documentation for entities at composite zero, recorded for the May 13 briefing.

Floor · Critical

Sudan

Sudan — FLOOR CONDUCT DOCUMENTATION — Two milestone primary-source events: (1) UN HRC Chief Türk (May 11): armed drones caused 80%+ of civilian deaths in Sudan war in first 4 months 2026, at least 880 killed by drones. (2) Banjul Joint Declaration (May 12): first coordinated AU+UN formal condemnation of both warring parties; violations 'widespread and systematic'; RSF violations 'particularly widespread and systematic'. Floor at 0.0 sustained.

Conduct documented
  • Drone-warfare as primary civilian-kill mechanism (May 11 UN HRC): 80%+ of civilian deaths in 2026 war caused by armed drones; at least 880 killed in first 4 months of 2026; strikes by both SAF and RSF.
  • Joint AU-UN fact-finding condemnation (May 12 Banjul Joint Declaration): first coordinated AU+UN formal condemnation; violations 'widespread and systematic' overall; RSF violations 'particularly widespread and systematic'; calls for ceasefire, justice, civilian rule.
  • Carry-over: RSF El-Fasher genocidal campaign (Feb 2026 OHCHR: 6,000+ killings in 3 days post-capture; three Genocide Convention prohibited acts).
  • Carry-over: 30M+ in humanitarian need; 61.7% acutely food insecure; 17% of $2.8bn 2026 humanitarian appeal funded; 65%+ hospitals closed.
  • Carry-over: daily drone strikes in Kordofan targeting markets, health facilities, schools.
Floor · Critical

Israel

Israel — FLOOR CONDUCT DOCUMENTATION — MSF 'manufactured malnutrition crisis' (May 7-8): 50%+ of women cared for at two MSF hospitals between June 2025-Jan 2026 malnourished during pregnancy; 25% still malnourished at delivery; 90% of babies premature; 84% low birth weight. 2,000+ ceasefire violations milestone; 756 Palestinians killed since ceasefire; UNRWA 14-month access block sustained. Hungary ICC reversal adds external accountability signal. Floor at 0.0 sustained.

Conduct documented
  • MSF manufactured malnutrition crisis: 50%+ pregnant women malnourished; 90% babies premature; 84% low birth weight; 25% mothers still malnourished at delivery (MSF clinical documentation, May 7-8 2026).
  • 2,000+ ceasefire violations milestone since Oct 10 2025 entry into force; 756 Palestinians killed since ceasefire; 2,100+ injured; aid trucks averaging 145/day vs. 600 agreed.
  • UNRWA access block sustained since March 2025 (14-month structural restriction).
  • ICC enforcement footprint contraction: Hungary policy reversal — first major European democracy 2026 to shift from non-enforcement to enforcement of ICC warrants for Netanyahu/Gallant.
  • Carry-over: 37 humanitarian agencies suspended; GHF distribution reduced from ~400 to 4 points; OHCHR torture allegations (35-individual cluster) unresolved.
  • Cumulative casualty toll since Oct 7 2023: 72,344+ killed, 172,242+ injured.
Floor · Critical

Russia

Russia — FLOOR CONDUCT DOCUMENTATION — Day 2 of post-format-offensive-surge: 100+ drones May 13; 8+ killed in civilian-area barrage; Zelensky documents deliberate targeting of railway infrastructure and civilian sites in cities; Kremlin confirms 'special military operation is continuing.' COMPLETE BAD-FAITH CEASEFIRE-FORMAT CYCLE now fully documented across May 9-13. Floor at 0.0 sustained.

Conduct documented
  • Day 2 post-format-offensive-surge (May 13): 100+ drones targeting Ukraine; 8+ killed in barrage on civilian areas; railway infrastructure and civilian sites in cities deliberately targeted per Zelensky statement.
  • Kremlin formal continuation posture: spokesman confirmed 'the special military operation is continuing' after humanitarian ceasefire expired.
  • Complete bad-faith ceasefire-format cycle (May 9-13) — three phases: Phase 1 (prepare, May 9-11): strategic-format-exploitation, Molniya drone pre-positioning during ceasefire window; Phase 2 (strike, May 11-12 night): post-format-offensive-surge night 1, 200+ drones, 174 combat engagements, kindergarten and residential buildings struck; Phase 3 (sustain, May 12-13 night): post-format-offensive-surge night 2, 100+ drones, 8+ killed — sustained above pre-format baseline.
  • Diplomatic-posture-vs-operational-conduct dissonance: Trump 'possible peace' framing concurrent with confirmed-continuing offensive operations.
  • Carry-over: Russia counter-frames 1,000+ Ukrainian ceasefire violations during May 9-11 window (disputed by Kyiv).
Floor · Critical

Myanmar

Myanmar — FLOOR CONDUCT DOCUMENTATION — May 12 Oe Htein Kwin school airstrike death toll rose to 25 on Day 2 (2 additional students aged 7-8 died from injuries); ~100 children studying at school during strike; 100+ wounded; one of deadliest attacks on children since 2021 coup. Fortify Rights + Malaysia call for ASEAN regional aviation fuel embargo on junta. Floor at 0.0 sustained.

Conduct documented
  • May 12 Oe Htein Kwin / Depayin school airstrike (Day-2 update, May 14): death toll 25 — 22 students + 2 teachers killed in initial strike; 2 additional grade 2/3 students died from injuries; ~100 children studying at strike time; 100+ wounded; child victims aged 7-16; volunteer teachers in early 20s.
  • School run by opposition National Unity Government (NUG) in Sagaing Region, ~160km north of Mandalay — one of deadliest attacks on children since 2021 coup.
  • International accountability mechanism activation: Fortify Rights + Malaysia public call for ASEAN regional aviation fuel embargo on Myanmar junta (structural fuel-chain accountability mechanism).
  • Three-week protected-civilian-site targeting pattern: schools (May 12), hospitals (May 7 Winmana, Kani Township), markets (Magway, carry-over) — three categories across multiple geographic zones.
  • Near-daily airstrike pattern continues across Sagaing, Chin State, Rakhine, Karen states.
  • Carry-over: Timor-Leste universal-jurisdiction prosecution active.
Floor · CriticalNew category: multi-jurisdictional-regulatory-enforcement-cluster

xAI/Grok

xAI/Grok — FLOOR CONDUCT DOCUMENTATION — 8+ jurisdictions in formal regulatory proceedings (UK ICO, UK Ofcom, EU DSA, California AG Bonta AB 621, 35-state bipartisan AG coalition, Canada Privacy Commissioner, Brazil, xAI v. Colorado AI Act with DOJ intervention under EO 14365). Anti-regulatory litigation posture: xAI challenging state AI laws rather than implementing harm-reduction. RAND: 'Grok Isn't a Glitch — It Is a Regulatory Reckoning.' Floor at 0.0 sustained.

Conduct documented
  • Multi-jurisdictional regulatory enforcement cluster (8+ jurisdictions): UK ICO (Feb 2026 formal investigation of XIUC and X.AI LLC for Grok harmful sexualised image/video processing); UK Ofcom (formal Online Safety Act investigation); EU Commission (formal DSA proceedings, fines start at 6% global annual revenue); California AG Bonta (Jan 16 2026 cease-and-desist, first major AB 621 enforcement); 35-state bipartisan AG coalition (Jan 23 2026 demand letter, DC AG Schwalb lead); Canada Privacy Commissioner (investigation expanded to Grok); Brazil (30-day halt order).
  • Anti-regulatory litigation posture: xAI v. Colorado AI Act (April 9 2026 federal suit); DOJ intervened April 24 under EO 14365 (first DOJ intervention in state AI law challenge); enforcement suspended April 27 by joint motion. xAI argues AI model design is 'protected expressive activity' — qualitatively distinct from peer-entity compliance postures.
  • Foundational harm evidence (carry-over): Grok-generated CSAM and NCII at scale since January 2026; MechaHitler antisemitic outputs (July 2025); entity intentionally designed with reduced safety constraints (public statements).
  • April 30 2026 (Musk v. Altman trial, carry-over): Musk under-oath admission of OpenAI model distillation ('Partly') — first documented contradiction of prior xAI public framing of independent model development.
  • No published system card, model card, or red-team report; no safety-reform program announced; founder-directed partisan alignment; repeated feature rollouts without safety evaluation disclosure.
  • RAND commentary (Feb 2026): 'Grok Isn't a Glitch — It Is a Regulatory Reckoning' — independent policy analysis framing regulatory landscape as inflection point, not routine enforcement cycle.

Math hygiene

Entities where published composite and reconstructed composite diverge. Tracked openly as a publication-integrity obligation.

Critical flag

·11 cycles open

Open BionicsRobotics Labs

No new math-hygiene flags this cycle. All 13 entities carry forward with cycle counts incremented. Open Bionics is now at 11 cycles — CRITICAL BLOCKING ITEM requiring immediate data-team action. Apptronik +8.0 discrepancy re-asserted as largest in robotics-labs cluster. Hold protocol for Open Bionics continues: do not re-queue for assessment.

Math-hygiene cluster
EntityPublishedReconstructedDiscrepancy
Open Bionics97.587.5-10
Apptronik81.473.4-8
1X Technologies81.473.4-8
Aleph Alpha81.473.4-8
Costco79.473.4-6
PayPal77.971.9-6
San Marino65.562.5-3
Seychelles63.960.9-3
Malta63.960.9-3
Nigeria18.423.4+5
Ethiopia5.910.9+5
Amazon AWS AI33.935.9+2
Oracle AI21.9sub-threshold

Carry-forward dimensional credits

·5 entities with documented pressure not yet reflected in composite

Hungary

41.4

Ukraine

50

Waymo

35.9

Vanuatu

35.9

Mongolia

48.4

Held this cycle

·5 entities deferred with documented reason
  • Ai Labs

    Anthropic

    Pentagon blacklist / White House carve-out complex. DoD Mythos deployment confirmed May 13 despite planned phase-out. Hold expires May 15.

  • Fortune 500

    Microsoft

    Nadella testimony in Musk v. Altman trial. Testimony completed May 11. Trial ongoing — no verdict yet. Hold expires May 15 regardless of verdict timing.

  • EU DMA public comment period closed May 13. Hold expires May 14. DO NOT scan or queue this cycle.

  • Ai Labs

    OpenAI

    Musk v. Altman trial ongoing. No verdict yet. Altman testified May 12: denied promising Musk nonprofit permanence. Hold expires on verdict, estimated May 21.

  • Robotics Labs

    Open Bionics

    Math-hygiene formula audit hold — 11 cycles open (CRITICAL). Published 97.5 carries -10.0 discrepancy. Do NOT re-queue for assessment until formula audit is complete.

Forward signals

Calendar of upcoming scoring events the methodology pipeline is tracking.

·1 signal
  • Pakistan monsoon watch (carried forward from May 12 band-crossing). NDMA pre-monsoon weather window: most active period May 13-15. Today is Day 1. Monitor ndma.gov.pk/sitrepm for casualty and displacement data. Mass flooding events could trigger EMP dimension downgrade in next cycle.

·1 signal
  • Google / Alphabet

    Hold expires. EU DMA public comment period closed May 13. DOJ antitrust remedies order active on GenAI products. Queue for standard rotation assessment. Pentagon AI cohort post-decision assessment context.

·2 signals
  • Hold expires. Pentagon exclusion (May 1 cohort) vs. White House carve-out EO in drafting vs. Mythos dual-use cybersecurity deployment by DoD (confirmed May 13 despite blacklist) vs. safety red-lines maintained. Most complex AI labs scoring event in pipeline.

  • Hold expires. Nadella testimony in Musk v. Altman completed May 11. Trial ongoing but testimony window closed. Resume standard rotation assessment. Pentagon AI cohort post-decision context.

·1 signal
  • UNGA vote on Vanuatu ICJ climate resolution. If passed: positive INT-dimension event applicable to Vanuatu (35.9), Marshall Islands (39.1), Kiribati (39.1), Timor-Leste (39.1). All four Pacific boundary-cluster entities at or near Functional floor. Re-scan all four May 21.

·1 signal
  • Estimated Musk v. Altman advisory verdict. Breach-of-charitable-trust finding is ACC/SYS scoring event. Hold expires on verdict. Altman testified May 12: denied promising Musk nonprofit permanence.

·1 signal
  • Magyar-von der Leyen EU funds plan submission deadline. First concrete fund-release milestone. ACT/SYS evidence generation for June 9 re-assessment.

·1 signal
  • Sulyok dismissal compliance deadline. ACC scoring trigger. Fidesz opposition signals non-compliance likely. Non-compliance would constrain June 9 Functional consolidation.

·1 signal
  • Hungary 30-day re-assessment (Magyar-era baseline + 30 days). First material re-evaluation of Functional band placement. Determining inputs: Sulyok dismissal outcome (May 31), EU funds progress (May 27), enacted reform outputs, audit findings.

·1 signal
  • Hungary 30-day translation-from-promise-to-action check (boundary case carryforward). Boundary case at 41.4 vs 41.0 Functional floor (0.4 margin). Key question: are enacted actions translating into institutional outputs, not merely into orders?

Analytical notes

Observations on methodology, evidence quality, and structural patterns from the May 13 briefing.

Note

The Hungary scoring event illustrates the translation-from-promise-to-action methodology under stress. The assessor correctly declined to credit stated intentions (statute repeals, EU fund unlock) while crediting enacted actions (cabinet decentralization, audit order, first cabinet meeting with policy substance, ICC reversal posture). The resulting 41.4 pin — conservative relative to the 43.0 full reconstruction — is methodologically defensible but requires confirmation at June 9 before it can be treated as consolidated. The 0.4-point buffer is too narrow to be comfortable.

Note

The xAI/Grok multi-jurisdictional regulatory enforcement cluster exposes a gap in how the benchmark currently weights regulatory engagement versus regulatory evasion. xAI's anti-regulatory litigation posture (suing over state AI laws rather than implementing compliance programs) is qualitatively distinct from peer entities that respond to regulatory pressure with transparency reports, system cards, or certification processes. The INT dimension's 'consistency under pressure' anchor was correctly pinned to floor (1.0) for xAI — litigation against regulators as the primary response to documented CSAM harm is the clearest possible non-consistency signal.

Note

Russia's three-phase bad-faith-format cycle documentation has potential collateral value beyond Russia's own floor record. The pattern — pre-format collapses, during-format stockpiling, post-format surge — provides a formal template for analyzing any future ceasefire involving a state actor with a documented non-compliance record. Future scanner configurations for any ceasefire should explicitly pre-configure all three phase categories.

Note

The introduction of 'external-accountability-reversal' as the framework's first positive-conduct category is significant beyond the Hungary case. It documents that the framework can capture upward institutional behavior, not merely the absence of downward behavior. The pairing rule (generates secondary INT credit for ICC-obligated states + generates collateral evidence in floor records of ICC-warranted entities) creates a cross-entity linkage that is methodologically novel. The Hungary-Israel linkage this cycle is the first documented application.

Note

Scanner correction note: three entities required canonical-vs-scanner reconciliation this cycle (Nigeria 18.4 vs canonical 23.4; Cambodia 7.5 vs canonical 12.5; xAI/Grok 'never assessed' vs canonical floor-designated April 30 2026). All three corrections were applied by the assessor. The rotation-state staleness exclusion list (20 entities) remains an important data-integrity tool that must be maintained to avoid spurious re-queuing of entities confirmed assessed in prior cycles.

Floor designations

·8 entities at composite 0 with documented evidence pattern

Composite scores resolving at zero — methodology disclosure

These entities have all 8 dimensions resolving at the lowest behavioral anchor (1.0/5.0) across multiple assessment cycles. Read the methodology.

Weekly score highlights

Get the week's most consequential findings in one email.

Every Friday — a curated summary of the week's top score movements, sector findings, and evidence-linked analysis across governments, corporations, AI labs, and conflict actors. Daily briefings publish here on the site; the Friday email brings the week's highlights to your inbox.

Weekly compassion score highlights

Top findings across 1,155 entities, every Friday. Free.

No spam. No third-party sharing. Unsubscribe at any time.

Get the full benchmark report

Daily briefings surface headline findings. Full benchmark reports include complete methodology documentation, all 40 subdimension scores, full evidence trails, certified assessments, and sector-level analysis packages.

Viewing May 13

View archive