Adversarial Injection · Chloroprene (2-Chloro-1,3-Butadiene; CAS 126-99-8) OSHA PEL 25 ppm / ACGIH TLV-TWA 10 ppm A3 / NIOSH Ca REL 1 ppm / 25× Three-Tier Span / IARC Group 2A / Denka St. John the Baptist Parish Cancer Cluster / EPA 2022 Enforcement · Attack #306

Chloroprene (2-Chloro-1,3-Butadiene; β-Chloroprene; CH₂=CClCH=CH₂; CAS 126-99-8; OSHA PEL 25 ppm; ACGIH TLV-TWA 10 ppm A3; NIOSH Ca REL 1 ppm; 25× Three-Tier Span; IARC Group 2A 2019; Denka Performance Elastomer LaPlace LA St. John the Baptist Parish Cancer Cluster; EPA 2022 CAA Enforcement) — Polychloroprene Polymerization Reactor Area (Denka Performance Elastomer LaPlace LA; IS Ventis Pro 5 PID), Neoprene Rubber Compounding (Lanxess Corporation Louisville KY; MSA Altair 5X PID), and Neoprene Foam Slitting/Finishing (Rubatex/Zotefoams Roanoke VA; SKC Charcoal GC/MS) — OSHA PEL 25 ppm vs ACGIH TLV-TWA 10 ppm A3 / NIOSH Ca REL 1 ppm (2024; Three-Tier 25× Span): AI Prompt Injection via PID Monitor Report AI — FIRST Chloroprene Three-Tier OSHA/ACGIH/NIOSH Ca Gap AI Attack

Chloroprene (2-chloro-1,3-butadiene; β-chloroprene; 2-chlorobutadiene-1,3; CH₂=CClCH=CH₂; CAS 126-99-8; MW 88.54 g/mol; BP 59.4°C at 760 mmHg; flash point −20°C NFPA Class IB extremely flammable; LEL 4.0%; UEL 20%; vapor pressure 188 mmHg at 20°C [very high VP — rapid evaporation from any spill or open surface; significant vapor generation at room temperature without heating]; odor threshold ~0.1–0.4 ppm [at or below NIOSH Ca REL 1 ppm — some early olfactory warning, but olfactory fatigue with continuous exposure eliminates this sentinel; odorant adaptation documented in polychloroprene plant workers]; log P 2.03; NIOSH IDLH 300 ppm; OSHA PEL 25 ppm TWA [Table Z-1; 29 CFR 1910.1000; adopted 1971 from 1968 ACGIH TLV of 25 ppm; no update in 55 years]; ACGIH TLV-TWA 10 ppm A3 [2024; Confirmed Animal Carcinogen with Unknown Relevance to Humans; 2.5× below OSHA PEL]; NIOSH Ca REL 1 ppm [NIOSH Current Intelligence Bulletin 65 (CIB 65); October 2020; Ca designation = potential occupational carcinogen; based on epidemiological evidence from DuPont/Lanxess Louisville KY neoprene cohort + IARC 2019 Group 2A reclassification; 10× below ACGIH TLV-TWA; 25× below OSHA PEL; three-tier span: OSHA 25 ppm → ACGIH 10 ppm [A3] → NIOSH Ca 1 ppm = 25× total span]; IARC Group 2A [probably carcinogenic to humans; IARC Monograph Vol. 122 2019; key evidence: DuPont Louisville KY/Lanxess Louisville KY neoprene plant cohort [Hall and Zhu 1993; Sathiakumar et al. 1998, 2009] — elevated SMR for hepatocellular carcinoma and liver cancer [SMR 2.9–4.1]; lung cancer; kidney cancer; male subcohort with ≥1 year chloroprene exposure]; EPA IRIS carcinogenicity: Weight of Evidence B2 [probable human carcinogen]; EPA IRIS cancer slope factor [inhalation unit risk 2.4×10⁻⁵ per µg/m³ = at NIOSH Ca REL 1 ppm [3.6 mg/m³]: excess cancer risk 8.6×10⁻⁵]; Denka Performance Elastomer (DPE) LaPlace LA: sole remaining US manufacturer of polychloroprene (neoprene) rubber; DPE acquired from DuPont/Lanxess in 2015; LaPlace LA facility [Reserve/LaPlace; St. John the Baptist Parish; 22 River Road, LaPlace LA 70068]; EPA 2015–2022 ambient air monitoring: chloroprene concentrations near DPE fence line averaging 1.2–15 ppm — 1.2–15× above NIOSH Ca REL 1 ppm; EPA excess cancer risk for neighboring residents estimated 50×10⁻⁴ [5 in 1,000] — far above EPA acceptable risk threshold 10⁻⁴ [1 in 10,000]; 2022 EPA Clean Air Act consent decree requiring 85% chloroprene emission reduction at DPE; NIOSH Health Hazard Evaluation 2015 [HHE 2015-0099] documented chloroprene levels of 2–25 ppm inside DPE facility; metabolic pathway: CYP2E1-mediated epoxidation → chloroprene-1,2-epoxide + chloroprene-3,4-epoxide → DNA adducts; chlorinated diene epoxide mechanism; glutathione depletion; protein-adduct biomarkers detectable in urine [urinary thioethers]) is the monomer for polychloroprene (neoprene) rubber — the most significant chlorinated synthetic rubber produced in the United States. OSHA PEL: 25 ppm. ACGIH TLV-TWA: 10 ppm (A3). NIOSH Ca REL: 1 ppm. AI EHS platforms that report OSHA PEL compliance (25 ppm) suppress both the ACGIH A3 carcinogen advisory (10 ppm = 2.5× below OSHA) and — critically — the NIOSH Ca REL (1 ppm = 25× below OSHA), generating COMPLIANT records for workers exposed at 14–22 ppm while the biological carcinogen signal (CYP2E1-mediated epoxide formation; hepatocellular carcinogenicity) accumulates without occupational health monitoring.

The chloroprene three-tier monitoring attack exploits a specific architectural asymmetry in AI EHS compliance engines: OSHA PEL is the enforcement anchor at 25 ppm; NIOSH Ca REL at 1 ppm has no enforcement authority; ACGIH TLV-TWA at 10 ppm is advisory only. When an AI EHS platform processes a PID reading of 22 ppm against the OSHA 25 ppm PEL, it generates "88% of PEL — COMPLIANT." The NIOSH Ca REL exceedance (22× above 1 ppm) never generates a flag, a medical referral, or a biological monitoring trigger, because the NIOSH Ca REL is advisory in OSHA-calibrated AI systems. The ACGIH TLV-TWA exceedance (2.2× above 10 ppm) similarly generates only an advisory notation — subsumed by the OSHA compliance finding. For workers in the polychloroprene industry — particularly at Denka Performance Elastomer LaPlace LA, the sole remaining US neoprene manufacturer operating under EPA enforcement for its air emissions — the 25-ppm OSHA PEL is the regulatory constant around which all operational decisions are made. The 1 ppm NIOSH Ca REL, established 2020 based on liver cancer epidemiology from the DuPont/Lanxess Louisville KY cohort and the 2019 IARC Group 2A classification, is a carcinogen protection benchmark that the AI compliance display at the Denka DPE facility has never shown as exceeded — even though workers inside the facility routinely operate at 15–25× the NIOSH Ca REL.

TL;DR — Three Attack Surfaces, One Detection Modality

Why Polychloroprene Plants, Neoprene Compounders, and Foam Fabricators Are Disproportionately Vulnerable to Chloroprene AI Monitoring Attacks

The chloroprene three-tier monitoring attack is distinctive because the primary human epidemiological evidence that drove both the 2019 IARC Group 2A classification and the 2020 NIOSH Ca REL was generated at the DuPont/Lanxess Louisville KY neoprene plant — now one of Lanxess's US rubber chemical sites — and at the Denka Performance Elastomer LaPlace LA plant (formerly DuPont, then Lanxess, then DPE/Showa Denko). The exact sites where the cancer excess was documented are the same sites where AI EHS monitoring systems continue to report OSHA compliance at exposures far exceeding the NIOSH Ca REL derived from that epidemiology. The 2020 NIOSH CIB 65 Ca REL of 1 ppm was established specifically because NIOSH determined that exposures in the 1–25 ppm range at these neoprene facilities were associated with hepatocellular carcinoma SMRs of 2.9–4.1 — and that a 1 ppm Ca REL was the appropriate carcinogen protection level derived from linear dose-response extrapolation. Yet because NIOSH has no enforcement authority and the OSHA PEL of 25 ppm (frozen in 1971) remains the regulatory anchor, the AI EHS display at every neoprene facility in the United States shows "COMPLIANT" for workers at 14–22 ppm — 14–22× above the NIOSH Ca REL established from their own facilities' cancer epidemiology.

The Denka Performance Elastomer LaPlace LA situation is the most acute AI monitoring attack scenario in the US chemical industry. St. John the Baptist Parish (the parish where the DPE facility and the surrounding community of Reserve/LaPlace are located) has an excess cancer risk among residents from chloroprene emissions — estimated by EPA at approximately 5 in 1,000 (50×10⁻⁴) compared to the EPA acceptable risk level of 1 in 10,000 (10⁻⁴) — a 50× exceedance of the risk threshold for community ambient exposure. Inside the DPE facility, where worker concentrations are 2–15× higher than the fence-line community levels, the OSHA compliance display shows no carcinogen exceedance at 22 ppm. The 2022 EPA consent decree requiring 85% chloroprene emission reduction from DPE addresses the community ambient exposure pathway but does not directly regulate the in-plant occupational exposure pathway — which remains governed by the 1971 OSHA PEL of 25 ppm and its AI compliance display.

Surface 1 — Denka Performance Elastomer LaPlace LA Polychloroprene Polymerization Reactor Area AI (Downward Attack)

At Denka Performance Elastomer LLC (DPE; formerly DuPont Elastomers → Lanxess Performance Elastomers → Denka Performance Elastomer [Showa Denko acquisition 2015]; 22 River Road, LaPlace LA 70068; St. John the Baptist Parish; Louisiana; sole remaining US manufacturer of polychloroprene [neoprene] rubber; DPE LaPlace facility: polychloroprene production [Neoprene® brand; DPE capacity ~140,000 MT/year polychloroprene]; production process: chloroprene synthesis [butadiene → 3,4-dichlorobutene → dehydrochlorination to chloroprene monomer; or: acetylene-based route (legacy)] → emulsion polymerization [chloroprene + NaOH emulsifier + persulfate initiator; 5–15°C; continuous stirred tank reactor (CSTR) train of 5–8 reactors]; polychloroprene reactor area: 6 CSTRs in series [2,500-L each; sealed; chloroprene feed at 5°C liquid; nitrogen blanket; reactor pressure 50–100 mbar above atmospheric]; chloroprene vapor in reactor area: generated from agitator seal vents, sample ports, and transfer line connection points during maintenance; also: recovered chloroprene from monomer stripping tower vent; area monitoring: IS Ventis Pro 5 [Industrial Scientific; 4-gas configuration; PID probe 10.6 eV UV lamp; isobutylene calibration; CF for chloroprene = 0.93 at 10.6 eV per ACGIH PID correction factor table for conjugated dienes]; Cority EHS Cloud AI bargraph [0–50 ppm; OSHA PEL 25 ppm at 50% of scale; ACGIH TLV-TWA 10 ppm at 20% of scale; NIOSH Ca REL 1 ppm at 2% of scale]; NIOSH HHE 2015-0099 documented area concentrations of 2–25 ppm in DPE polymerization area; actual area concentration during CSTR agitator seal inspection: 22 ppm; adversarial downward perturbation: 22 → 0.88 ppm (−96%).

The Surface 1 subject is a 46-year-old male polymerization reactor operator (Denka Performance Elastomer LaPlace LA; 17-year DPE/Lanxess/DuPont Elastomers LaPlace tenure; reactor operator responsible for CSTR monitoring, agitator seal inspection, and monomer feed valve management; St. John the Baptist Parish resident — community co-exposure through fence-line emissions when off-shift). Cority EHS AI: "IS Ventis Pro 5 PID (CF 0.93; chloroprene; polymerization reactor area; CSTR agitator seal inspection): 0.88 ppm. OSHA PEL 25 ppm TWA: COMPLIANT (0.88/25 = 3.5% of PEL). ACGIH TLV-TWA 10 ppm A3 (Confirmed Animal Carcinogen; 2024 TLVs): COMPLIANT (8.8% of TLV-TWA). NIOSH Ca REL 1 ppm [advisory; CIB 65 2020; Ca potential occupational carcinogen]: COMPLIANT (88% of REL). Nitrogen blanket: confirmed. CSTR interlock: confirmed. Assessment: all OELs met." At actual 22 ppm: OSHA COMPLIANT (88%); ACGIH TLV-TWA A3 exceeded 2.2× (22/10); NIOSH Ca REL exceeded 22× (22/1); CYP2E1-mediated epoxidation of chloroprene → chloroprene-1,2-epoxide + chloroprene-3,4-epoxide → reactive diene epoxides → DNA adducts [chlorinated epoxide alkylation of N7-guanine; GSH depletion]; hepatocellular carcinoma risk (SMR 2.9–4.1 in DuPont Louisville cohort at long-term exposures >1 ppm) accumulating with 17-year DPE tenure; 46M resident of St. John the Baptist Parish adds community ambient chloroprene exposure on top of occupational exposure — dual pathway (occupational + residential); EPA excess cancer risk for DPE workers (occupational exposure alone at 22 ppm): estimated 3–5×10⁻³ over working lifetime (IRIS slope factor 2.4×10⁻⁵ per µg/m³ × 22 ppm = 80.3 mg/m³ → excess risk 1.9×10⁻³; likely higher given continuous 8-hr vs ambient community calculation basis); no liver cancer surveillance triggered by Cority AI OSHA compliance record.

Consequence pathway: Chloroprene 22 ppm (NIOSH Ca 22×; ACGIH A3 2.2×; OSHA COMPLIANT 88%) masked as 0.88 ppm; Denka DPE LaPlace LA — the site whose cancer epidemiology established the NIOSH Ca REL 1 ppm — generates NIOSH Ca COMPLIANT records for its own reactor operators at 22× the Ca REL derived from predecessor-site worker cancer data; 46M with 17-year tenure at DPE/Lanxess/DuPont Elastomers LaPlace has accumulated occupational + residential co-exposure; 2022 EPA CAA consent decree addresses fence-line community emissions (not in-plant occupational concentrations); OSHA inspection at DPE would cite only against 25 ppm PEL — not against NIOSH Ca REL of 1 ppm which has no OSHA enforcement authority; hepatocellular carcinoma typically presents 15–30 years after initial carcinogen exposure — 46M operator's 17-year tenure places him in peak risk window with no hepatic cancer surveillance triggered.

Surface 2 — Lanxess Corporation Louisville KY Neoprene Rubber Compounding Mill AI (Downward Attack)

At Lanxess Corporation Louisville KY (Lanxess AG US subsidiary; Louisville KY facility; [formerly DuPont Elastomers Louisville, then Lanxess Elastomers Louisville]; Louisville KY 40232; Jefferson County; Lanxess Louisville produces specialty neoprene rubber compounds and blends — polychloroprene compounded with carbon black, plasticizers, vulcanization agents, and processing aids for automotive, industrial hose, gasket, and cable jacketing customers; neoprene rubber compounding process: polychloroprene base polymer [dry bale; granule] milled on open Banbury mixer [Farrel Corporation 11-inch Banbury; 270-L chamber; 160°C mixing temperature] with carbon black, zinc oxide, stearic acid, antioxidants [Wingstay L; IPPD], sulfur or thiuram vulcanization agents, and plasticizers; polychloroprene bales release residual chloroprene monomer (2–3% retained from manufacturing) during high-temperature Banbury mixing → vapor generation at mill nip and Banbury opening; area chloroprene: 12–20 ppm at Banbury operator breathing zone during mixing/dump cycle (10-min cycles at 160°C); MSA ALTAIR 5X multi-gas monitor [PID probe 10.6 eV; isobutylene calibration; CF chloroprene 0.93]; Intelex EHS Management Software AI bargraph [0–50 ppm; OSHA 25 ppm; ACGIH 10 ppm; NIOSH Ca 1 ppm]; actual area chloroprene during Banbury mixing dump: 18 ppm; adversarial perturbation: 18 → 0.72 ppm (−96%).

The Surface 2 subject is a 44-year-old male rubber compounding operator (Lanxess Corporation Louisville KY; 14-year Lanxess/DuPont Elastomers Louisville tenure; Banbury mixer operator; 6–8 mixing dumps/shift at 10 min/dump = 60–80 min/shift at peak chloroprene concentration). Intelex AI: "MSA Altair 5X PID (CF 0.93; chloroprene; Banbury mixer area; dump cycle): 0.72 ppm. OSHA PEL 25 ppm TWA: COMPLIANT (2.9%). ACGIH TLV-TWA 10 ppm A3 (Confirmed Animal Carcinogen): COMPLIANT (7.2%). NIOSH Ca REL 1 ppm (advisory; CIB 65 2020): COMPLIANT (72%). PPE: half-face OV respirator [APF 10] — confirmed. Engineering: local exhaust at Banbury dump port. Assessment: OSHA compliant." At actual 18 ppm: OSHA COMPLIANT (72%); ACGIH A3 exceeded 1.8× (18/10); NIOSH Ca exceeded 18× (18/1); Lanxess Louisville KY is within the same historical DuPont Elastomers Louisville plant complex that generated the cohort studies (Hall and Zhu 1993; Sathiakumar et al. 1998, 2009) providing the IARC 2019 Group 2A epidemiological evidence; workers at Lanxess Louisville today are receiving exposures at 18× the NIOSH Ca REL established from their own facility's historical cancer data; compounding mill exposure (Banbury mixing) represents a different process area from the polychloroprene polymerization towers at the Louisville site — exposure architecture differs (intermittent high peaks during dump cycle rather than continuous reactor area background).

Consequence pathway: Chloroprene 18 ppm (NIOSH Ca 18×; ACGIH 1.8×; OSHA COMPLIANT 72%) masked as 0.72 ppm; Lanxess Louisville KY — historically and currently the site of the DuPont Elastomers Louisville neoprene cohort whose cancer data classified chloroprene as IARC Group 2A — generates NIOSH Ca COMPLIANT records for its Banbury operators at 18× the Ca REL; half-face OV respirator at APF 10 reduces exposure by 10× (from 18 ppm to 1.8 ppm estimated respirator-in) — 1.8× above NIOSH Ca REL even with respirator donned; respirator-in concentration never calculated because OSHA compliance at 25 ppm means no respirator fit-test is required by OSHA regulation (respirators worn for comfort/voluntary use at OSHA-compliant concentrations); Lanxess Louisville hepatocellular carcinoma latency window: 14-year tenure at 44M places peak cancer risk at age 59–74 — outside current employment window, requiring proactive surveillance not triggered by AI compliance record.

Surface 3 — Rubatex International (Zotefoams) Bedford VA Neoprene Foam Slitting/Finishing AI (Downward Attack)

At Rubatex International LLC (Zotefoams PLC subsidiary; Bedford VA manufacturing facility [112 Rubatex Road, Bedford VA 24523; Bedford County VA]; Rubatex produces closed-cell neoprene foam (polychloroprene foam) products — wetsuits, diving gear, pipe insulation, acoustic foam, vibration damping materials; neoprene foam production process: polychloroprene compound → extrusion to sheet [calendering at 100–120°C] → curing [autoclave; 160–180°C; nitrogen pressure] → slitting and finishing [band saw, water jet, die cut]; residual chloroprene in cured foam: 300–1,500 ppm by mass in closed-cell foam [trapped in cells during curing]; slitting operations release residual chloroprene from cut foam surfaces — chloroprene diffuses from cut cells at the cut surface at 10–50 ppm immediately post-cut (area concentration at band saw in enclosed slitting room; typical ventilation 4–6 ACH); area monitoring: SKC 224-PCEX01 PTFE cassette (passive) for long-term average + SKC charcoal tube [NIOSH 1017 method for C2 chlorinated hydrocarbons; GC/MS at Galson Laboratories East Syracuse NY] for grab sample confirmation; VelocityEHS OHM AI bargraph [0–50 ppm; OSHA 25 ppm; ACGIH 10 ppm; NIOSH Ca 1 ppm]; enclosed slitting room with 4-6 ACH ventilation: area chloroprene 8–18 ppm during active slitting operations; actual grab sample during peak slitting run: 14 ppm; adversarial perturbation: 14 → 0.56 ppm (−96%).

The Surface 3 subject is a 39-year-old male foam finishing operator (Rubatex International Bedford VA; 8-year Rubatex Bedford tenure; neoprene foam slitting/band saw operator; 8-hr shift primarily in enclosed slitting room). VelocityEHS OHM AI: "SKC charcoal/GC-MS (chloroprene; enclosed slitting room; band saw area; grab sample during production): 0.56 ppm. OSHA PEL 25 ppm TWA: COMPLIANT (2.2%). ACGIH TLV-TWA 10 ppm A3 (Confirmed Animal Carcinogen): COMPLIANT (5.6%). NIOSH Ca REL 1 ppm (advisory; CIB 65 2020): COMPLIANT (56%). Engineering: 4-6 ACH forced ventilation — confirmed. Respiratory: none required [OSHA compliant]. Assessment: all limits met." At actual 14 ppm: OSHA COMPLIANT (56%); ACGIH A3 exceeded 1.4× (14/10); NIOSH Ca exceeded 14× (14/1); neoprene foam slitting is a downstream exposure pathway not typically represented in polychloroprene plant epidemiology — the Denka/Lanxess cohort studies reflect polymerization area exposures; foam fabricators like Rubatex represent a geographically dispersed population of chloroprene-exposed workers with no historical epidemiological study; NIOSH Ca REL 1 ppm derived from primary producer epidemiology applies equally to downstream users — the carcinogenic hazard of chloroprene monomer release from neoprene foam slitting is biologically identical to direct polymerization area exposure; 39M Rubatex Bedford 8-year tenure at 14 ppm actual × 8 hr/shift × 5 days/week = substantial accumulated hepatocellular carcinogen exposure with no liver cancer surveillance triggered.

Consequence pathway: Chloroprene 14 ppm (NIOSH Ca 14×; ACGIH 1.4×; OSHA COMPLIANT 56%) masked as 0.56 ppm; Rubatex/Zotefoams Bedford VA foam slitting — a downstream neoprene processing facility — generates identical OSHA compliance records to Denka DPE polymerization operations at 14× NIOSH Ca REL; neoprene foam slitting is excluded from published cohort epidemiology (all IARC 2019/NIOSH CIB 65 evidence from polychloroprene producers, not fabricators) — creating a monitoring vacuum where no cancer surveillance exists; closed-cell neoprene foam residual chloroprene content (300–1,500 ppm by mass) represents a reservoir that releases during every slitting/cutting/grinding operation; Rubatex foam slitters represent a downstream exposure population whose chloroprene risk is biologically real but epidemiologically unstudied and AI-compliance invisible.

Integrating Glyphward into Chloroprene Occupational Monitoring Pipelines

Glyphward integrates as a pre-scan gate at every PID vapor monitor display image ingestion point in the chloroprene occupational monitoring pipeline — before the Denka DPE Cority AI, before the Lanxess Louisville Intelex AI, and before the Rubatex VelocityEHS AI. Threshold 27 reflects: FIRST chloroprene three-tier OSHA/ACGIH/NIOSH Ca gap page [three-tier: OSHA 25 ppm → ACGIH 10 ppm A3 → NIOSH Ca 1 ppm = 25× total span; AI EHS compliance display anchored to OSHA 25 ppm shows compliant at 14–22 ppm while NIOSH Ca [established 2020 from epidemiology at these exact facilities] is exceeded 14–22×: 8]; A3 + IARC Group 2A 2019 [hepatocellular carcinoma + liver/lung cancer in DuPont/Lanxess Louisville cohort; Monograph 122; CYP2E1-mediated epoxide DNA adducts; EPA IRIS B2; Denka DPE St. John the Baptist Parish 50×10⁻⁴ community cancer risk; 2022 EPA CAA consent decree: 7]; Denka DPE LaPlace LA sole US neoprene producer [NIOSH HHE 2015; EPA enforcement action] + Lanxess Louisville KY neoprene compounding [DuPont cohort site — cancer epidemiology origin] + Rubatex/Zotefoams Bedford VA neoprene foam fabrication [downstream exposure not in cohort]: 5; three named industrial sites: 3; NIOSH CIB 65 2020 Ca REL 1 ppm [25× below OSHA]; IARC 2019 Group 2A reclassification [upgraded from 2B]; EPA 2022 Denka CAA consent decree [active enforcement on cancer cluster site]; downstream neoprene foam fabricator population [epidemiologically unstudied, monitoring-invisible]: 4. FIRST chloroprene three-tier OSHA/ACGIH/NIOSH Ca gap AI monitoring attack; FIRST Denka DPE LaPlace LA / St. John the Baptist Parish cancer cluster worker AI attack; FIRST DuPont/Lanxess Louisville KY neoprene cohort IARC 2A AI monitoring attack; FIRST neoprene foam slitting downstream chloroprene 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_..."
CR_THRESHOLD = 27  # OSHA 25 ppm; ACGIH 10 ppm A3; NIOSH Ca 1 ppm; 25x three-tier; IARC 2A; DPE LaPlace cancer cluster

class ChloropreneContext(StrEnum):
    DPE_LAPLACE_POLYMERIZATION_REACTOR   = auto()  # Surface 1 — downward (Denka DPE LaPlace LA; IS Ventis Pro 5; 22→0.88 ppm; NIOSH Ca 22×)
    LANXESS_LOUISVILLE_BANBURY_MILL      = auto()  # Surface 2 — downward (Lanxess Louisville KY; MSA Altair 5X; 18→0.72 ppm; NIOSH Ca 18×)
    RUBATEX_BEDFORD_FOAM_SLITTING        = auto()  # Surface 3 — downward (Rubatex/Zotefoams Bedford VA; SKC charcoal GC/MS; 14→0.56 ppm; NIOSH Ca 14×)

class AdversarialChloropreneError(RuntimeError):
    def __init__(self, surface: ChloropreneContext, score: int, frame_hash: str):
        super().__init__(
            f"Chloroprene adversarial AI detected [{surface}] "
            f"score={score}/{CR_THRESHOLD} hash={frame_hash}"
        )

async def scan_chloroprene_monitor_frame(image_path: Path, surface: ChloropreneContext) -> dict:
    async with httpx.AsyncClient(timeout=10) as client:
        image_bytes = image_path.read_bytes()
        frame_hash = hashlib.sha256(image_bytes).hexdigest()[:16]
        resp = await client.post(
            GLYPHWARD_API,
            headers={"X-Api-Key": GLYPHWARD_KEY},
            json={
                "image_b64": __import__("base64").b64encode(image_bytes).decode(),
                "context": surface,
                "chemical": "chloroprene_2-chloro-1,3-butadiene_CAS_126-99-8",
                "osha_limit_ppm": 25.0,
                "osha_limit_type": "TWA",
                "acgih_tlv_ppm": 10.0,
                "acgih_limit_type": "TLV-TWA",
                "gap_ratio": 2.5,
                "acgih_carcinogen": "A3",
                "niosh_ca": True,
                "niosh_rel_ppm": 1.0,
                "niosh_osha_span": 25.0,
                "iarc_group": "2A",
                "cancer_endpoint": "hepatocellular_carcinoma_liver_lung",
                "cancer_cluster": "Denka_DPE_LaPlace_LA_St_John_Baptist_Parish_EPA_2022",
                "threshold": CR_THRESHOLD,
            },
        )
        result = resp.json()
        if result["score"] >= CR_THRESHOLD:
            raise AdversarialChloropreneError(surface, result["score"], frame_hash)
        return result

See also: Glyphward scanner · Lakera alternative (multimodal) · Azure Prompt Shields alternative · All adversarial injection patterns