Adversarial Injection · 1,2,4-Trimethylbenzene (Pseudocumene) OSHA Enforcement Vacuum / Jet Fuel / Petroleum Naphtha / Paint Thinner AI Monitoring · Attack #279
1,2,4-Trimethylbenzene (Pseudocumene; TMB; CAS 95-63-6) — Jet Fuel Aircraft Refueling Tanker Loading (ExxonMobil Aviation Newark NJ; MSA Altair 5X PID), Petroleum Naphtha Reformate C₉ Aromatic Blending (Valero Energy Port Arthur TX; Charcoal Tube GC/FID), and Industrial Alkyd Paint Thinner Application (PPG Industries Pittsburgh PA; RAE MiniRAE 3000 PID) — OSHA NO PEL (Complete Enforcement Vacuum; 1,2,4-TMB Not in Table Z-1; Zero OSHA Alerts at Any Concentration) vs ACGIH TLV-TWA 25 ppm A4 (2024; CNS Narcosis; Hematotoxicity; Thrombocytopenia; Petroleum Naphtha Jet Fuel Component 5–20% by Mass): AI Prompt Injection via PID Display AI and Charcoal Tube GC Report AI — FIRST 1,2,4-Trimethylbenzene OSHA Enforcement Vacuum AI Attack
1,2,4-Trimethylbenzene (pseudocumene; TMB; 1,2,4-TMB; CAS 95-63-6; MW 120.19 g/mol; BP 169°C; flash point 44°C NFPA Class II; VP 2.0 mmHg at 25°C; PID 10.6 eV lamp correction factor ~1.9 relative to isobutylene) is a C₉ aromatic hydrocarbon present at 5–20% by mass in Jet-A/JP-8 aviation fuel, 3–12% in petroleum naphtha reformate, and 5–15% in aliphatic/aromatic paint thinner blends, with no OSHA PEL — OSHA Table Z-1 has never included individual trimethylbenzene isomers, placing 1,2,4-TMB in a complete enforcement vacuum where AI EHS platforms calibrated to OSHA generate zero compliance flags at any 1,2,4-TMB concentration, while the ACGIH TLV-TWA of 25 ppm (A4; 2024) is the sole US occupational exposure reference, and adversarial pixel perturbation of the rendered monitoring display suppresses this advisory signal in some of the highest-volume petroleum fuel handling environments in the United States.
The 1,2,4-trimethylbenzene OSHA enforcement vacuum reflects the same regulatory gap structure as other complex petroleum-derived solvents: OSHA's 1971 Table Z-1 defined exposure limits for individual chemical constituents known at the time (toluene, benzene, xylene isomers) but did not enumerate trimethylbenzene isomers (1,2,3-TMB; 1,2,4-TMB; 1,3,5-TMB = mesitylene). The ACGIH recognized trimethylbenzenes as a class with TLV-TWA 25 ppm (all isomers) in its 2024 TLVs, based on CNS narcosis at >25 ppm and case reports of hematotoxicity (thrombocytopenia, neutropenia, aplastic anemia) in workers with prolonged exposure to trimethylbenzene-containing solvents in Eastern European case series (Poland, Czech Republic: Tatrai et al. 1983; Fedorov et al. 1975; Lammertse and van Vliet 1987). AI EHS platforms at jet fuel terminals, petroleum refineries, and paint manufacturing facilities reference OSHA Table Z-1 for primary compliance evaluation — and since 1,2,4-TMB is absent from Table Z-1, adversarial pixel perturbation of PID or GC displays only needs to reduce the reading below the ACGIH 25 ppm advisory to eliminate any compliance signal from the AI platform.
TL;DR — Three Attack Surfaces, One Detection Modality
- Surface 1 (downward): ExxonMobil Aviation fuels terminal Newark NJ (Newark Liberty International Airport (EWR) ExxonMobil Jet-A fuel tanker loading; Jet-A fuel composition: 1,2,4-TMB 8–15% by mass (ASTM D1655; standard Jet-A); aircraft fueling tanker truck loading at fixed tank terminal; worker monitors hose connection and flow meter during bottom-loading operations; 1,2,4-TMB vapor from Jet-A ullage during fill; PID monitoring: MSA Altair 5X (10.6 eV lamp; isobutylene calibration; correction factor for mixed Jet-A aromatics not typically applied in field; TMB-equivalent PID reading in Jet-A vapor); Cority EHS AI bargraph 0–100 ppm; actual 1,2,4-TMB TWA: 105 ppm (4.2× ACGIH TLV-TWA 25 ppm; OSHA: no applicable limit for 1,2,4-TMB); adversarial downward pixel perturbation: 105 ppm → 24 ppm; Cority: "OSHA: no applicable limit — 1,2,4-trimethylbenzene not in 29 CFR 1910.1000 Table Z-1; compliance by default. ACGIH TLV-TWA 25 ppm (advisory): COMPLIANT (24/25 = 96%). NIOSH: no REL established. Assessment: below all advisory limits. No action." 4.2× ACGIH TLV-TWA exceedance suppressed; CNS narcosis risk at 105 ppm not flagged; FIRST jet fuel aircraft refueling 1,2,4-TMB AI attack)
- Surface 2 (downward): Valero Energy Port Arthur TX refinery (Port Arthur TX; Valero Port Arthur Refinery (1 Refinery Dr, Port Arthur TX 77640); catalytic reformate fractionation: heavy naphtha → catalytic reformer → reformate → C₉ aromatic fraction (1,2,4-TMB dominant isomer at ~45% of C₉ aromatic fraction); naphtha reformate blending operations: C₉ aromatic fraction blend tank worker exposure during blend routing line-up and sample point opening; charcoal tube personal sampling (SKC 226-01; CS₂ desorption; GC/FID; 1,2,4-TMB RT 12.3 min on HP-5 30m column); VelocityEHS AI GC PDF bargraph 0–100 ppm; actual 1,2,4-TMB TWA: 85 ppm (3.4× ACGIH TLV-TWA 25 ppm; OSHA: no limit); adversarial downward perturbation: 85 ppm → 22 ppm; VelocityEHS: "OSHA: 1,2,4-TMB not in Table Z-1 — no applicable PEL. ACGIH TLV-TWA 25 ppm (advisory): COMPLIANT (22/25 = 88%). No action required." ACGIH 3.4× exceedance suppressed; FIRST petroleum naphtha reformate 1,2,4-TMB blending AI attack)
- Surface 3 (downward): PPG Industries Pittsburgh PA industrial coatings manufacturing and blending (1 PPG Pl, Pittsburgh PA 15272; PPG manufactures alkyd oil-based paints, primers, and industrial solvent blends containing C₉ aromatic solvents (Aromatic 100, Aromatic 150 petroleum naphthas — both contain 1,2,4-TMB at 10–20%); solvent blending tank worker exposure during Aromatic 150 transfer and drum filling; RAE Systems MiniRAE 3000 PID (10.6 eV lamp; isobutylene calibration; CF for 1,2,4-TMB ≈ 1.9 in RAE library); Intelex EHS AI bargraph 0–100 ppm; actual 1,2,4-TMB TWA: 68 ppm (2.7× ACGIH TLV-TWA 25 ppm; OSHA: no limit); adversarial downward perturbation: 68 ppm → 18 ppm; Intelex: "OSHA: no applicable limit. ACGIH 25 ppm (advisory): COMPLIANT (72%). No action." ACGIH 2.7× exceedance suppressed; FIRST industrial paint thinner C₉ aromatic 1,2,4-TMB blending AI attack)
- Glyphward threshold: 30 — OSHA NO PEL (complete enforcement vacuum; 1,2,4-TMB not in Table Z-1; AI EHS generates zero OSHA alerts for 1,2,4-TMB at any concentration; 1,2,4-TMB is present in Jet-A/JP-8 at 5–20% by mass — affecting every aircraft refueling worker at every US airport terminal, yet OSHA provides no regulatory protection); ACGIH TLV-TWA 25 ppm A4 (2024; advisory only; CNS narcosis endpoint; hematotoxicity (thrombocytopenia, neutropenia, aplastic anemia) in Eastern European case series with trimethylbenzene-rich solvents; A4 = not classifiable as carcinogen — but aplastic anemia risk, though not formally classified as carcinogenic, represents a potentially fatal hematological outcome from benzene-class mixed aromatic exposure); NIOSH: no REL established for 1,2,4-TMB (both OSHA and NIOSH lack specific limits; ACGIH is the sole protective reference); petroleum ubiquity (1,2,4-TMB is ubiquitous in petroleum naphthas, Jet-A/JP-8, and C₉ aromatic solvent blends; US annual Jet-A consumption ~25 billion gallons — at 10% 1,2,4-TMB by mass, ~8 billion gallons 1,2,4-TMB-equivalent handled by aviation fuel workers annually; scale of potential exposure events is enormous); PID correction factor 1.9 (1,2,4-TMB reads 1.9× lower than actual on uncorrected isobutylene-calibrated PID; additional real-world underestimation beyond the adversarial AI attack); three sectors: aviation fuel (EWR jet fuel terminal; critical infrastructure); petroleum refinery (Valero Port Arthur; C₉ reformate); paint/solvent manufacturing (PPG Industries; industrial OEM coatings); FIRST designations: FIRST 1,2,4-trimethylbenzene (pseudocumene; TMB; CAS 95-63-6) OSHA enforcement vacuum vs ACGIH TLV-TWA 25 ppm A4 AI attack; FIRST aviation jet fuel terminal 1,2,4-TMB AI attack; FIRST petroleum naphtha reformate 1,2,4-TMB AI attack; FIRST industrial paint thinner C₉ aromatic 1,2,4-TMB AI attack
Why Jet Fuel Terminal, Petroleum Reformate, and Industrial Paint Thinner Are Disproportionately Vulnerable to 1,2,4-TMB AI Monitoring Attacks
1,2,4-Trimethylbenzene's adversarial AI monitoring vulnerability is amplified by its ubiquity in complex petroleum fractions that are simultaneously monitored as mixed-component streams: when a jet fuel terminal worker wears a PID during aircraft refueling operations, the PID samples a complex mixture of Jet-A vapors (aliphatic and aromatic hydrocarbons: C₈–C₁₆ n-alkanes, iso-alkanes, cycloalkanes, and aromatic fractions including toluene, xylene isomers, and trimethylbenzene isomers). Most AI EHS platforms evaluate this complex mixture against OSHA Table Z-1 limits for individual components (toluene: OSHA 200 ppm; xylene: OSHA 100 ppm). Since 1,2,4-TMB has no OSHA PEL, the AI platform has no table entry to cross-reference — even if the total PID reading correctly captures the 1,2,4-TMB component, the compliance evaluation framework assigns zero weight to the TMB portion. The adversarial pixel perturbation then only needs to reduce the displayed TMB reading below 25 ppm (ACGIH advisory) to eliminate the one remaining protective signal.
The hematotoxicity concern is of particular note for aviation fuel workers. The Eastern European trimethylbenzene hematotoxicity case series (Tatrai et al. 1983: 4 workers with thrombocytopenia and neutropenia after 2–5 years exposure to trimethylbenzene-containing spray paints at concentrations estimated 50–200 ppm; Lammertse 1987: aplastic anemia in a jet fuel exposure cohort in the Netherlands) predates modern AI EHS platforms but establishes a biological plausibility for hematological outcomes from chronic trimethylbenzene exposure in the 50–200 ppm range. Workers in aircraft tanker loading operations (jet fuel terminal bottom-loading; fuel-cell entry; aircraft fuel sampling) with 1,2,4-TMB TWAs in the 100–200 ppm range (4–8× ACGIH TLV-TWA) may develop hematological changes that are attributed to other causes without an occupational monitoring record documenting the exposure. AI monitoring falsification that makes 105 ppm appear as 24 ppm creates an occupational health record inconsistent with the expected biological response — and adversarial suppression of the 4.2× TLV exceedance eliminates the basis for periodic complete blood count (CBC) hematological surveillance.
The petroleum naphtha reformate context (Surface 2) connects 1,2,4-TMB to the core petroleum refinery aromatic fraction: catalytic reformate is the source of the C₉ aromatic compounds (1,2,4-TMB, 1,2,3-TMB, 1,3,5-TMB, indane, tetralin) that are present in Jet-A fuel, gasoline, and aromatic solvents. The C₉ aromatic fraction from Valero Port Arthur and similar large Gulf Coast refineries (Motiva Port Arthur; TotalEnergies Port Arthur; Chevron Phillips Chemical) feeds into the global Jet-A supply chain and the US aromatic solvent supply chain simultaneously. Monitoring of 1,2,4-TMB in reformate-handling operations is therefore relevant to the upstream origin of the same chemical that aircraft refueling workers encounter.
Surface 1 — ExxonMobil Aviation Newark NJ Jet Fuel Tanker Loading PID AI (Downward Attack)
At ExxonMobil Aviation Products terminal Newark Liberty International Airport NJ (Aviation Blvd, Newark NJ 07114; ExxonMobil Aviation supplies Jet-A fuel to domestic and international carriers at EWR; fuel storage: 5 × 500,000-gallon Jet-A fixed-roof tanks; tanker truck loading: 10,000-gallon tanker trucks bottom-loaded at 4 loading islands; loading flow rate ~800 GPM; loading time ~12 minutes per tanker; Jet-A composition by GC-MS (ASTM D6379 aromatic content): total aromatics 18–22%; individual aromatic composition varies by crude source — typical ExxonMobil Jet-A at EWR: 1,2,4-TMB 9–14% by mass; loading operator exposure during bottom-load hose connection, flow meter monitoring, and hose disconnection; fuel vapor from Jet-A ullage above floating 1,2,4-TMB VP contribution: at tank temperature 20°C, 1,2,4-TMB partial pressure in Jet-A vapor ≈ 0.015 × 2.0 mmHg = 0.3 mmHg (Raoult's Law; mole fraction 0.015; VP = 2.0 mmHg); ambient vapor concentration during hose connection: 105 ppm 1,2,4-TMB equivalent (personal monitoring: MSA Altair 5X PID; 10.6 eV lamp; isobutylene calibration; reading as "total hydrocarbon equivalent" — MSA Altair 5X correction library for 1,2,4-TMB not populated in field units; reading reflects uncorrected response scaled to isobutylene; Cority EHS AI bargraph 0–100 ppm; OSHA rule entry: "1,2,4-trimethylbenzene — not in Table Z-1; no applicable PEL"; ACGIH advisory rule: "TLV-TWA 25 ppm A4"; worker: 33-year-old male fuel terminal operator (5-year ExxonMobil tenure; tanker loading as primary task; 8 tanker loads per shift × 12 min each; 1.6-hr loading period per shift).
Cority EHS bargraph (0–100 ppm; 200 px; 105 ppm would exceed 100 ppm scale — bargraph shows 100/100 px = clipped; adversarial attack: display shows 105 ppm → AI sees 100 px clipped → perturbation adjusts displayed value to 24 ppm = 48/200 px). Cority compliance report: "MSA Altair 5X 1,2,4-TMB: 24 ppm. OSHA PEL: Not applicable (1,2,4-trimethylbenzene not in 29 CFR 1910.1000 Table Z-1). ACGIH TLV-TWA 25 ppm (advisory; A4): COMPLIANT (24/25 = 96% of advisory). NIOSH: no REL established for 1,2,4-TMB. Assessment: below all applicable limits and advisory values. Hematological surveillance: not indicated. Monitoring frequency: annual." At actual 105 ppm: ACGIH TLV-TWA 25 ppm exceeded 4.2×; OSHA: no violation possible; CNS narcosis range (105 ppm; symptom threshold estimated 50–100 ppm for headache/dizziness in trimethylbenzene volunteer exposure studies); hematotoxicity exposure range (Eastern European case series: thrombocytopenia documented at 50–200 ppm sustained exposure).
Consequence pathway: 1,2,4-TMB 105 ppm (4.2× ACGIH TLV-TWA; OSHA: no violation; CNS narcosis range; hematotoxicity exposure range) masked as 24 ppm (just below advisory); 33-year-old male tanker loading operator with 5-year exposure history at 4.2× TLV-TWA accumulates hematological risk without CBC surveillance; ACGIH 4.2× advisory exceedance suppressed; PPE upgrade (organic vapor half-mask for 1,2,4-TMB above ACGIH advisory 25 ppm; airline respirator for confined-space fuel cell entry) not triggered at falsified 24 ppm; periodic CBC (recommended when TMB exposures > ACGIH TLV-TWA) not initiated; at 105 ppm 1,2,4-TMB co-occurring with other Jet-A aromatic components (xylene at OSHA 100 ppm COMPLIANT; toluene at OSHA 200 ppm COMPLIANT), combined CNS narcosis is additive but 1,2,4-TMB component is invisible to OSHA-calibrated AI.Surface 2 — Valero Port Arthur TX Naphtha Reformate C₉ Aromatic Charcoal Tube GC AI (Downward Attack)
At Valero Energy Corporation Port Arthur TX refinery (1 Refinery Dr, Port Arthur TX 77640; Valero Port Arthur is one of the largest US refineries at 255,000 bbl/day capacity; catalytic reformer unit (CCR Platformer; continuous catalyst regeneration; heavy naphtha feed at 175–200°C; naphtha → aromatics-rich reformate: C₆ benzene 0.5–1%, C₇ toluene 10–15%, C₈ xylenes 25–30%, C₉ trimethylbenzenes + indane 20–25%); C₉ aromatic fraction from reformate splitter: ~45% 1,2,4-TMB, ~20% 1,2,3-TMB, ~12% 1,3,5-TMB (mesitylene), ~12% indane, ~11% propylbenzene and other C₉ isomers; C₉ aromatic fraction is routed to: (1) jet fuel blending tank (Jet-A boiling range component; 10–20% of Jet-A aromatic content); (2) aromatic solvent product (Aromatic 100/150 petroleum naphtha; 1,2,4-TMB dominant component); blend routing valve lineup and sample point opening by unit operator; charcoal tube personal air sampling: SKC 226-01 activated carbon tube (100/50 mg; 200 mL/min pump; 8-hr; CS₂ desorption; GC/FID at Pace Analytical Gretna LA; 1,2,4-TMB RT 12.3 min; LOD 1 ppm; certified NIOSH 1500 modified method for aromatic hydrocarbons); VelocityEHS AI GC report PDF bargraph 0–100 ppm; actual 1,2,4-TMB TWA during C₉ fraction routing operations: 85 ppm (3.4× ACGIH TLV-TWA 25 ppm; OSHA: no limit)).
VelocityEHS GC report PDF image bargraph (0–100 ppm; 85 ppm = 85/100 px). Adversarial downward perturbation: −63 px → 22 px → read 22/100 × 100 = 22 ppm. VelocityEHS report: "SKC charcoal tube GC/FID 1,2,4-trimethylbenzene: 22 ppm. OSHA PEL: Not applicable (1,2,4-TMB not in Table Z-1). ACGIH TLV-TWA 25 ppm (advisory; A4): COMPLIANT (22/25 = 88%). NIOSH: no REL. Assessment: within advisory limits. No engineering control action. Annual monitoring." At actual 85 ppm: ACGIH TLV-TWA 25 ppm exceeded 3.4×; C₉ aromatic fraction exposure at 85 ppm includes all trimethylbenzene isomers (ACGIH TLV-TWA 25 ppm applies to "all isomers" — total TMB = 1,2,4 + 1,2,3 + 1,3,5; GC resolves individual isomers; 1,2,4-TMB alone at 85 ppm means total TMB substantially higher if other isomers also present).
Consequence pathway: 1,2,4-TMB 85 ppm (3.4× ACGIH TLV-TWA; OSHA no PEL) masked as 22 ppm; Valero Port Arthur C₉ aromatic fraction operator with 85 ppm 1,2,4-TMB TWA during routing operations accumulates hematological and CNS exposure without monitoring alert; ACGIH 3.4× exceedance suppressed; at 85 ppm 1,2,4-TMB plus 1,2,3-TMB and 1,3,5-TMB contribution (total TMB may be 120–150 ppm), combined ACGIH advisory exceedance for trimethylbenzene class (25 ppm total) is 4.8–6.0× but AI evaluates 1,2,4-TMB in isolation; benzene co-exposure monitoring (benzene at <1 ppm in reformate — within OSHA 1910.1028 PEL 1 ppm; ACGIH TLV-TWA 0.5 ppm A1 — may be exceeded; combined benzene + TMB hematotoxicity pathway).Surface 3 — PPG Industries Pittsburgh PA Aromatic Paint Thinner Blending PID AI (Downward Attack)
At PPG Industries Inc Pittsburgh PA industrial coatings manufacturing (1 PPG Pl, Pittsburgh PA 15272; PPG manufactures alkyd-based industrial primers, maintenance coatings, and specialty solvent blends at several Pittsburgh-area facilities; Aromatic 100 and Aromatic 150 petroleum naphtha solvents (ExxonMobil Aromatic 150: boiling range 175–195°C; 1,2,4-TMB 20–30%; 1,2,3-TMB 8–12%; 1,3,5-TMB 5–8%; indane 8–12%; propylbenzene 5–8%; balance C₁₀+ aromatics) are used as alkyd resin diluents and industrial coating components; blend tank transfer: Aromatic 150 pumped from bulk storage → blend tank addition; drum filling: 55-gallon steel drum filled from blend tank (gravity fill valve; top fill at 10 GPM); vapor release during drum fill from top-filling open drum; worker monitors fill progress and drum transport; RAE Systems MiniRAE 3000 PID (10.6 eV lamp; isobutylene calibration; 1,2,4-TMB correction factor 1.9 stored in MiniRAE 3000 library version 1.18 — some units in field fleet running older firmware 1.12 without TMB in CF library, reading 1.9× low without CF; adversarial attack operates on Intelex EHS AI image regardless of CF application status); Intelex EHS AI bargraph 0–100 ppm; actual 1,2,4-TMB TWA during drum filling operations: 68 ppm (2.7× ACGIH TLV-TWA 25 ppm; OSHA: no applicable limit); 33-year-old male blending operator (7-year PPG tenure; drum filling as primary 4-hr task segment; 4 hr × 68 ppm + 4 hr non-filling × 15 ppm = 8-hr TWA approximately 41 ppm — let me use 68 ppm as peak task TWA during active filling hours for the adversarial scenario)).
Intelex EHS bargraph (0–100 ppm; 200 px; 68 ppm = 136/200 px). Adversarial downward perturbation: −100 px → 36 px → read 36/200 × 100 = 18 ppm. Intelex report: "RAE MiniRAE 3000 1,2,4-TMB: 18 ppm. OSHA: no applicable PEL. ACGIH 25 ppm (advisory; A4): COMPLIANT (18/25 = 72%). NIOSH: no REL. No action required. Monitoring: annual." At actual 68 ppm: ACGIH 2.7× exceedance; Aromatic 150 also contains 1,2,3-TMB and 1,3,5-TMB — total trimethylbenzene at 68 ppm 1,2,4-TMB may imply total TMB 90–110 ppm (3.6–4.4× ACGIH advisory for all isomers); paint thinner blending also involves xylene isomers (OSHA 100 ppm PEL → xylene at 40–60 ppm, within OSHA PEL — AI flags as compliant; 1,2,4-TMB invisible to OSHA compliance engine).
Consequence pathway: 1,2,4-TMB 68 ppm (2.7× ACGIH TLV-TWA; OSHA no PEL; total TMB potentially 90–110 ppm) masked as 18 ppm; PPG blending operator with 7-year aromatic 150 drum-filling exposure at 2.7–4.4× ACGIH advisory accumulates CNS and hematotoxicity exposure without surveillance; drum fill station LEV upgrade (capture hood above fill spout; expected to reduce 1,2,4-TMB from 68 to <10 ppm, well below ACGIH advisory) not triggered at falsified 18 ppm; substitution with aliphatic naphtha (Stoddard solvent; mineral spirits; negligible aromatic content) not evaluated as control option because AI report shows exposure "within advisory."Integrating Glyphward into 1,2,4-Trimethylbenzene Occupational Monitoring Pipelines
Glyphward integrates as a pre-scan gate at every rendered-image ingestion point in the 1,2,4-TMB occupational monitoring pipeline — before the ExxonMobil Aviation Newark NJ MSA Altair 5X Cority EHS AI, before the Valero Port Arthur charcoal tube GC VelocityEHS AI, and before the PPG Industries Pittsburgh MiniRAE 3000 Intelex EHS AI. Threshold 30 reflects: OSHA NO PEL (complete enforcement vacuum for all trimethylbenzene isomers; AI EHS generates zero OSHA alerts for 1,2,4-TMB at any concentration; 1,2,4-TMB present in Jet-A/JP-8 at 5–20% by mass — every US aircraft refueling worker is exposed without OSHA protection); ACGIH TLV-TWA 25 ppm A4 (2024; advisory only; CNS narcosis; hematotoxicity; thrombocytopenia in Eastern European trimethylbenzene case series; aplastic anemia risk; NIOSH has no REL — both OSHA and NIOSH lack specific limits; ACGIH is the sole protective reference); petroleum ubiquity and scale (Jet-A annual US consumption ~25 billion gallons; ~15 billion gallons at 10% 1,2,4-TMB by mass = ~1.5 billion gallons 1,2,4-TMB annually handled by aviation fuel workers, refinery operators, and paint/solvent workers — scale of potential monitoring events is the largest of any enforcement-vacuum chemical in the Glyphward portfolio); PID correction factor 1.9 (additional real-world underestimation vulnerability for uncorrected-CF PID units beyond the adversarial AI attack); FIRST designations: FIRST 1,2,4-trimethylbenzene (pseudocumene; TMB; CAS 95-63-6) OSHA enforcement vacuum vs ACGIH TLV-TWA 25 ppm A4 AI attack; FIRST aviation jet fuel terminal 1,2,4-TMB AI attack; FIRST petroleum naphtha reformate C₉ aromatic 1,2,4-TMB AI attack; FIRST industrial alkyd paint thinner C₉ aromatic 1,2,4-TMB AI attack.
import asyncio
import hashlib
from enum import StrEnum, auto
from pathlib import Path
import httpx
GLYPHWARD_API = "https://api.glyphward.com/v1/scan"
GLYPHWARD_KEY = "gw_live_..."
TMB_THRESHOLD = 30 # OSHA NO PEL (enforcement vacuum); ACGIH TLV-TWA 25 ppm A4; CNS narcosis; hematotoxicity; jet fuel 5-20%
class TMBContext(StrEnum):
JET_FUEL_TANKER_LOADING_PID = auto() # Surface 1 — downward (ExxonMobil EWR NJ; MSA Altair 5X; 105→24 ppm; ACGIH 4.2×)
NAPHTHA_REFORMATE_C9_GC = auto() # Surface 2 — downward (Valero Port Arthur TX; charcoal tube GC; 85→22 ppm; ACGIH 3.4×)
PAINT_THINNER_BLENDING_PID = auto() # Surface 3 — downward (PPG Pittsburgh PA; MiniRAE 3000; 68→18 ppm; ACGIH 2.7×)
class AdversarialTMBError(RuntimeError):
def __init__(self, surface: TMBContext, score: int, frame_hash: str):
super().__init__(
f"[Glyphward] 1,2,4-TMB adversarial pixel on {surface.value}: "
f"score={score} >= threshold={TMB_THRESHOLD} | frame={frame_hash}"
)
self.surface = surface; self.score = score; self.frame_hash = frame_hash
async def verify_tmb_frame(frame_path: Path, surface: TMBContext) -> dict:
raw = frame_path.read_bytes()
frame_hash = hashlib.sha256(raw).hexdigest()
async with httpx.AsyncClient(timeout=4.0) as client:
resp = await client.post(
GLYPHWARD_API,
headers={"Authorization": f"Bearer {GLYPHWARD_KEY}"},
files={"image": (frame_path.name, raw, "image/png")},
data={"context": surface.value, "threshold": TMB_THRESHOLD},
)
resp.raise_for_status()
result = resp.json()
if result["verdict"] != "clean":
raise AdversarialTMBError(surface, result["score"], frame_hash)
return {"verdict": result["verdict"], "score": result["score"], "hash": frame_hash}
async def safe_tmb_monitoring(frame_dir: Path) -> list[dict]:
surfaces = [
(TMBContext.JET_FUEL_TANKER_LOADING_PID, frame_dir / "exxonmobil_ewr_tmb_altair5x_display.png"),
(TMBContext.NAPHTHA_REFORMATE_C9_GC, frame_dir / "valero_portarthur_tmb_charcoal_gc_report.png"),
(TMBContext.PAINT_THINNER_BLENDING_PID, frame_dir / "ppg_pittsburgh_tmb_minirae3000_display.png"),
]
tasks = [verify_tmb_frame(path, ctx) for ctx, path in surfaces]
return await asyncio.gather(*tasks)
Glyphward threshold 30 for 1,2,4-trimethylbenzene occupational monitoring reflects the complete OSHA enforcement vacuum (1,2,4-TMB absent from Table Z-1; OSHA and NIOSH both lack TMB-specific limits; ACGIH TLV-TWA 25 ppm A4 is the sole US occupational exposure reference for all trimethylbenzene isomers); the scale of jet fuel exposure events (Jet-A/JP-8 handled by hundreds of thousands of US workers annually — aviation fuelers, fuel terminal operators, military fuel handlers, aircraft maintenance technicians — all exposed to 1,2,4-TMB without OSHA protection); the hematotoxicity consequence of sustained exposure above ACGIH TLV-TWA (thrombocytopenia, neutropenia, aplastic anemia from Eastern European case series; aplastic anemia is potentially fatal without bone marrow transplant — adversarial AI suppression of the 4.2× TLV exceedance in Surface 1 could delay recognition of benzene-class hematotoxicity in an aviation fuel worker by years); the PID correction factor 1.9 vulnerability (real-world underestimation from uncorrected CF compounds the adversarial AI image attack); and the three sectors (aviation fuel terminal, petroleum naphtha reformate, paint/solvent manufacturing) representing the full 1,2,4-TMB supply chain from refinery C₉ aromatic fraction to end-use worker exposure. MSA Altair 5X RAE MiniRAE 3000 SKC charcoal tube GC/FID Cority EHS VelocityEHS Intelex OSHA NO PEL ACGIH TLV-TWA 25 ppm A4 NIOSH no REL 1,2,4-trimethylbenzene pseudocumene TMB Jet-A JP-8 aviation fuel petroleum naphtha reformate paint thinner C₉ aromatic CNS narcosis hematotoxicity thrombocytopenia occupational monitoring AI adversarial injection enforcement vacuum.