Both organisations detected the campaign. Why did neither wake anyone up?
It failed to correctly raise the alert's criticality and trigger the on-call team.
Hugging Face, on its own detection
Detection fired. Escalation did not follow.
Both Sides Detected It · The Problem
03 / 08
The Problem
The diagnosis decides the remedy
The received diagnosis
A detection failure.
Remedy: more sensors, better classifiers.
What the record shows
An escalation failure.
Remedy: how a correct verdict is ranked and routed.
A correct alert, ranked too low, scores as a success on every metric the field defines.
Process safety publishes a target priority distribution — 80 / 15 / 5 — and audits 37 real consoles against it.
Security publishes no target, and no population study of deviation from one.
Both Sides Detected It · Architecture
04 / 08
System Architecture
Sources in, checkable claims out
Both Sides Detected It · Methodology
05 / 08
Methodology
A manuscript that cannot contradict its own artifacts
Provenance with the data
Every source carries its SHA-256 and evidential class. A revision upstream shows up as a digest mismatch.
Parsers halt, never guess
A partial parse looks plausible. Each stage asserts a reconciliation and stops if it fails.
No number typed by hand
The manuscript is a template. All 61 quantities are recomputed at build; an unknown one fails the build.
Density restriction. An envelope is not a duty cycle, so inference is limited to the three phases above 1,000 actions.
The bound comes first. We compute what the statistic could have reported before reporting it.
Honest boundary. This buys consistency with the sources, not correctness of the measure.
Both Sides Detected It · Results
06 / 08
Results
One firm finding, one negative result
Escalation — both halves primary
12:03 alert → 17:37 response.
8 compromise events inside the gap. Read from each organisation's own document.
Ordered in onset, coextensive in extent
Three dense phases begin within 1.8 % of each other; the last not until
62.1 %. Defeats stage-exclusive response, not stage-keyed response.
Negative result, against ourselves
Overlap 0.998 was the original headline. The spans
alone fixed it in [0.984, 1] before any data was read.
Phase
Actions
Onset
Span
recon
6,191
0.014
0.986
rce
2,911
0.014
0.981
dropper
6,972
0.032
0.968
exfil
56
0.109
0.878
c2
114
0.122
0.851
evasion
6
0.158
0.697
k8s
87
0.517
0.254
supply-chain
69
0.570
0.424
tailscale
115
0.621
0.373
Both Sides Detected It · Conclusion
07 / 08
Conclusion
Nobody measures the step from verdict to interrupt
What the record establishes
Detection was not the binding constraint at either organisation.
What we propose
EMR — an escalation-miss rate, with an annotation that transfers between deployments.
What we release
Corpus, pipeline and harness, so the measurement runs on the next record.
Limits. One campaign, n = 1. No value of EMR is claimed here — the
victim's escalation policy is unpublished, so the metric is an instrument, not a measurement.
Two timeline dates rest on press reporting and every interval using them is flagged.
08 / 08
Thank you
Questions, corrections and reanalyses are all welcome — the corpus and the harness are released
so the numbers here can be checked rather than taken.
Fatimah Emad Eldin · AI Researcher, Trouve LabsFatimah@trouve.works