Versioned editorial indicator
4SI / PHYSICAL TRUST INTELLIGENCE
State of Presence Risk 2026.
What official records reveal about the gap between digital claims and physical reality.
Safety Gate, CPSC and FDA
full editorial chains
deep institutional analysis
Digital proof can travel.
Reality does not.
Across product surveillance, human authority, agentic execution and infrastructure access, the recurring problem is not merely a false claim. It is the moment that claim is granted permission to change the physical or institutional world.
The published corpus contains 4,527 normalized records. Most are official product-safety records; 15 are separately researched Presence Incidents with a complete Claim → Authority → Action → Consequence chain. These populations are not interchangeable and are never presented as equally deep validation.
Four levels of
public evidence.
Each level answers a different question. A larger record count does not imply deeper validation.
4,512
3,978 EU Safety Gate, 524 U.S. CPSC and 10 FDA-hosted records normalized from official publication streams.
15
Source-validated cases where 4SI can map the complete analytical chain. This is not independent forensic verification.
10
Selected for institutional relevance, domain balance, source quality and explanatory depth.
4,527
Every record receives an ECR scenario range. Modeled exposure is not a documented loss sum.
Directional movement,
not measured prevalence.
The index is a versioned 4SI research position informed by evidence density, consequence, authority depth and verification difficulty. It is not a statistical estimator, probability or forecast.
INDEX MOVEMENT
52.0 to 86.0.
DOMAIN POSITION
Interpret with coverage.
4,518 of 4,527 records concern object presence. Domain scores must not be read as like-for-like prevalence comparisons.
Different surfaces.
The same control gap.
These are analytical patterns found across the official feeds and curated cases—not prevalence estimates for the wider economy.
A signal stands in for reality.
A mark, listing, credential, voice or message is accepted as sufficient proof of the person or object behind it.
Trust becomes permission.
The accepted claim enables payment, access, care, installation, custody, disclosure or execution.
The accountable source disappears.
Systems preserve a credential while losing the human, object or principal that made the credential meaningful.
Reality is checked too late.
The physical check happens after the consequential action, when recovery is slower, costlier or impossible.
Publication density reflects
the source system.
Counts show where the two official feeds published records, not comparative national risk.
France 572 · Italy 557 · United States 528 · Germany 487
EU values refer to notifying markets in Safety Gate; U.S. values primarily reflect CPSC publications. They should not be interpreted as incident rates.
Cosmetics 1,006 · Toys 739 · Electrical equipment 539
The record mix is dominated by the mandate and publication practices of product-safety authorities.
Chemical 1,514 · Injuries 646 · Electric shock 384
Labels reproduce normalized source taxonomy. A record may contain multiple risk types.
A modeled envelope,
not a loss total.
The full-corpus ECR is fixed at publication time. It combines documented monetary floors with modeled welfare, remediation, disruption and financial-flow exposure while suppressing likely overlap.
PUBLISHED PORTFOLIO ENVELOPE
$2.06B–$16.2B
$5.23BCentral modeled position · not a forecast
INTERPRETATION
The envelope is not a documented loss sum, exact loss prediction, reserve, valuation or portfolio-level P10/P50/P90 probability interval. Most records are proxy-led; correlation and possible record overlap are adjusted before aggregation, but cannot be eliminated from public evidence alone.
The boundary becomes clear
in concrete events.
Ten source-validated cases were selected for source quality, domain balance and institutional relevance.
Loading flagship cases
What this report
does not claim.
Public credibility depends on preserving the boundary between observed evidence and inference.
No prevalence claim
Source feeds reflect regulatory mandates and publication practices, not the full universe of Presence Risk.
No causal proof
4SI maps a control problem from published facts; it does not conduct independent forensic investigation.
No exact damages
ECR is a wide scenario envelope. Only the separately labeled floor is directly documented monetary impact.
No equal domain sample
Object records dominate the dataset. Other domains are represented through curated case analysis.
09 / INSTITUTIONAL MEANINGVerification must move
Verification must move
before execution.
Presence infrastructure connects the expected human, object or accountable principal to a consequential decision while intervention is still possible.