The most significant editorial findings in the Jul 29 briefing.
Editorial insight
Twenty governments, cities, companies and one university were reviewed overnight. Nineteen scores were confirmed within the benchmark's five-point range.
CMU's confidence cap rests on sourcing tier while Taipei's pending hold in the same queue rests on formula magnitude.
“CMU's proposed 5.5-point decline is capped at medium confidence because no tier-4 source exists; would a completed Title VI investigation outcome constitute the sourcing upgrade needed to resolve the pending queue's confidence cap?”
Tonight's review of Bangalore started from a claim that Human Rights Watch had documented new detentions of student protesters in Bengaluru around July 20 to 23, 2026. The benchmark's standard practice is to fetch every cited primary source in full before scoring it, not to trust a summary.
Why it matters
The evidence used to flag Bangalore for a protest crackdown never mentioned Bangalore. The city's real record this month was the opposite: an eight-day student protest it let continue peacefully.
Whether the WARN Act compliance investigations become filed complaints, which would supply the stronger source the current record lacks, is the next event to watch.
A coordinator-level comparison of Baxter's proposed reading against the Fortune-500 medical-device placeholder cohort, or a second Baxter recall, is the next event that would test this finding.
Any Ecuadorian or US government response to Human Rights Watch's inquiries, or a first prosecution arising from the documented incidents, is the next event to watch.
The response to the July 25 rally and any further prosecutions is the next event to watch.
How to read this briefing— Bands, scores, and terms
Schema guide
The 5 performance bands
critical0–20
developing20–40
functional40–60
established60–80
exemplary80–100
Two scales
Each of the 8 dimensions is scored 1.0–5.0; these combine into a 0–100 composite score, mapped to the 5 bands above.
Key terms
Band crossing
A score change large enough to move an entity from one performance band into an adjacent one — the most structurally significant finding in a given cycle.
Boundary watch
An entity whose current score is within 3 points of a band threshold, flagged for priority reassessment in the next cycle.
Carry-forward
A dimensional credit retained from a prior assessment when new evidence is insufficient to revise a specific dimension; disclosed explicitly.
First baseline
An entity's inaugural composite score — no published score exists to compare against, so delta is not shown.
Floor designation
The most serious finding: all 8 dimensions resolve at the lowest behavioral anchor (1.0/5.0) across multiple cycles, yielding a composite of 0.
Forward trigger
A scheduled future reassessment event (e.g., a policy implementation date or legislative deadline) that may materially change an entity's score.
Carnegie Mellon expanded discipline, including a possible expulsion, against students who painted its own designated free-speech wall. Its compassion score would fall 5.5 points, the only proposed change tonight.
Carnegie Mellon University maintains a physical wall on campus, known as the Fence, that the university itself designates for open student expression.
Where this sits
Read the full signal
Between July 13 and 29, 2026, CMU expanded formal Title VI and internal-conduct investigations of students who painted messages on the Fence from 8 people to 10, and campus-access restrictions from 7 to 9, with one student now facing possible expulsion. The university's own stated basis is that the messages constituted discriminatory harassment on the basis of race, national origin or religion. Some of the sanctioned messages were painted by Jewish students criticizing the Israeli government's policies, and nine organizations, including Jewish Voice for Peace and Palestine Legal, issued a joint statement on July 28, 2026 condemning the university's approach. Separately, on July 28, CMU revised its discrimination policy after receiving a poor rating from an outside advocacy group and described a disputed anti-Israel conduct clause in the revised text as a typographical error. A student petition opposing the crackdown had passed 400 signatures by July 29. The proposed 5.5-point decline is driven entirely by this conduct; the benchmark's formula bonus for broad institutional strength was already at zero in both the published and proposed scores, so none of the movement comes from a scoring mechanism. Confidence is capped at medium because the evidence is student journalism quoting the university directly, plus advocacy groups that are themselves party to the dispute; no independent government or court record exists yet. The published score of 49.2 is unchanged pending review.
Over time, a series of messages on the Fence crossed the bounds of our community norms, threatened to create a hostile environment for members of our Jewish community.
A citation used to question Milan's migrant-detention record was dated a year wrong. Once removed, the only real finding left was a housing clearance carried out with the city's own social workers present.
Tonight's review of Milan began with a claim that Italy's Constitutional Court had issued a July 2026 finding against migrant-detention practice. The primary source was fetched directly. The ruling is judgment No. 96, dated July 3, 2025, a full year before the claimed date.
Where this sits
Read the full signal
With that claim removed as out of window, only one piece of genuine in-window municipal conduct remained: on July 23, 2026, Milan's police, city police, waste services, the public housing agency and the city's own social services jointly cleared four occupied public-housing units in the Gratosoglio district, with social workers present throughout rather than notified afterward. Milan's score moved down slightly, well within normal review variance. Because the evidence base for Milan this cycle rests on a single local news item, confidence is set to low. This is the sixth consecutive night that a date or attribution error has been caught and corrected before it could reach a published score, following the same pattern as Bangalore tonight and Guayaquil, Atlanta and Mauritania in recent cycles.
A finding already built into Arizona's score nearly counted a second time. Three states that joined one lawsuit were checked individually instead of being treated as three separate findings from one event.
Arizona was flagged tonight on a July 17, 2026 court order finalizing a receiver over the state's prison health care. That order was already scored.
Where this sits
Read the full signal
Arizona's published score of 36.2 was set by its July 19, 2026 assessment, whose own findings already describe the same receiver and the same $500,000 salary. Scoring it again tonight would have counted the same failure twice, so the score held at 36.2. Separately, New Jersey, Minnesota and Colorado were all flagged on the same event: a 26-state lawsuit against FEMA and the Department of Homeland Security filed July 23, 2026. Joining a lawsuit signed by 25 other states is weak evidence about any one state, so each was checked on its own record instead. Minnesota had a genuine finding of its own: a federal court dismissed a case against its immigration-detainer law on July 21, 2026, after the state defended the law for ten months. New Jersey and Colorado had no comparable in-window conduct beyond the coalition filing. All three scores held unchanged.
Global Cities -- Two Citations, Two Different Kinds of Wrong
Bangalore was cited on a report about a different city; Milan was cited on a ruling dated a year wrong. Both corrections ran in the city's favor.
Read the full signal
Bangalore (32.8 published, 33.1 assessed): the triggering Human Rights Watch statement concerns New Delhi, not Bangalore; the verified Bengaluru record is a peaceful, week-long protest the city allowed to continue.
Milan (57.8 published, 56.9 assessed): the triggering Constitutional Court judgment is dated July 2025, not 2026; the sole in-window finding is a housing clearance conducted with the city's own social workers present.
Cairo (20.3 published, 20.6 assessed): nineteen of twenty detained Shia Muslims were remanded a further 15 days after interrogations that focused on their religious beliefs.
Phnom Penh (18.8 published, unchanged): scam-compound survivors freed by Cambodia's own crackdown are billed as visa overstayers rather than recognized as trafficking victims.
·medium
US Cities -- Federal Conduct Excluded on Both Sides
El Paso's worst findings were federal ICE conduct, not the city's; Los Angeles's homelessness count reversed two years of progress.
Read the full signal
El Paso (35.9 published, 35.6 assessed): a federal detention-contract extension and a regional surge in ICE arrests are not scored as city conduct under the attribution rule.
Los Angeles (35.9 published, 34.4 assessed): the official count found 45,194 people unhoused, a 3.4 percent rise, which the city's own homelessness authority links to a pause in new permanent-housing placements.
·medium
US States -- One Order Nearly Double-Counted, One Lawsuit Nearly Tripled
Arizona's prison-healthcare order was already priced into its score. Three states that joined one federal lawsuit were checked individually rather than counted as three findings.
Read the full signal
Arizona (36.2 published, unchanged): the July 17 receivership order was already reflected in the July 19 baseline; re-scoring it would have counted the same failure twice.
Minnesota (46.9 published, unchanged): a federal court upheld the state's refusal to enforce immigration detainers after a ten-month defense -- genuine own conduct, but one ruling is not an eight-dimension improvement.
New Jersey and Colorado (60.6 and 52.5 published, unchanged): both joined the same 26-state lawsuit against FEMA and DHS; coalition membership alone did not move either score.
11 signals shown
Risk signals
Developments that may affect future scores. Watch items from the Jul 29 briefing.
Risk
Carnegie Mellon's pending Title VI investigations, including one possible expulsion, could either produce the independent government record the current downgrade lacks or resolve without one.
Risk
Ten change proposals are now sitting in the queue with none applied in more than a week, spanning countries, cities, a university, an AI lab and two companies.
Risk
Milan's confidence is set to low because this cycle's entire in-window municipal-conduct record is a single local news item; a second independent source would materially change that.
Risk
Iran's public-execution pattern from the same mass trial has not stopped; a further execution would reinforce a score that is already near the floor without being able to move it lower.
Risk
Two Fortune 500 roster entries no longer reflect the companies they claim to track: Basic Energy Services was acquired in 2021 and Chesapeake Energy now operates as Expand Energy.
Score movements
Entities with score changes this cycle, followed by confirmed positions.
Further ECOWAS or CPLP action on detained political figures, or additional documentation of the transitional government's conduct, is the next event to watch.
A coordinator-level comparison of Baxter's proposed reading against the Fortune-500 medical-device placeholder cohort, or a second Baxter recall, is the next event that would test this finding.
Whether the WARN Act compliance investigations become filed complaints, which would supply the stronger source the current record lacks, is the next event to watch.
A coordinator-level review comparing Mali's conduct against Burkina Faso's (6.3 of 100) is the next scored event needed to resolve this gap. No review date has been set.
A coordinator-level review comparing Bolivia's conduct profile, as an elected government under economic and civil-unrest strain, against Critical-band peers facing state collapse or mass atrocity is the next event needed. No review date has been set.
methodology-evolution
Evidence ledger
Primary sources reviewed in this briefing cycle. 8 sources linked.
Human Rights Watch's July 23, 2026 statement on the student-protest crackdown documents events in New Delhi and does not mention Bengaluru or Bangalore anywhere in its text.
Student protesters continued a sustained demonstration at Bengaluru's Freedom Park for more than eight consecutive days through July 27, 2026, with volunteer-run water and food counters and no documented use of force.
Over time, a series of messages on the Fence crossed the bounds of our community norms, threatened to create a hostile environment for members of our Jewish community.
Nine organizations, including Jewish Voice for Peace and Palestine Legal, issued a joint statement on July 28, 2026 condemning CMU's handling of the Fence-painting investigations.
Italy's Constitutional Court judgment on detention in repatriation centers, cited as a 2026 finding, is dated July 3, 2025 -- a full year before the date claimed.
On July 23, 2026, Milan's police, city police, waste services, the public housing agency and the city's own social services jointly cleared four occupied public-housing units in the Gratosoglio district.
A federal judge finalized an order on July 17, 2026 installing a receiver, paid $500,000 a year, over health care for more than 25,000 people in Arizona's prisons; this order was already assessed on July 19, 2026.
A federal court dismissed a lawsuit challenging Minnesota's refusal to enforce federal immigration detainers on July 21, 2026, after the state defended the law for ten months.
Floor designations
·8 entities at composite 0 with documented evidence pattern
Composite scores resolving at zero — methodology disclosure
These entities consistently score the worst result across all 8 dimensions of compassionate conduct — the benchmark's most serious classification.
What “floor” means: every one of the 8 dimensions (Recognition, Response, Reduction, and 5 others) resolves at the lowest behavioral anchor (1.0/5.0) across multiple assessment cycles, yielding a composite score of 0. Full methodology.
Daily briefings surface the headline finding. Full benchmark reports include all 40 subdimension scores, complete evidence trails, certified assessments, and sector-level analysis packages — the record researchers and journalists cite.
Independence note: entities never pay for inclusion, score changes, or suppression of findings. Commercial services support access, interpretation, and institutional use only.