{
  "source": "Cooked Index — occupational AI risk register",
  "page": "https://cookedindex.com/jobs/textile-apparel-and-furnishings-workers-all-other/",
  "methodology": "https://cookedindex.com/methodology",
  "notice": "Verdicts are re-examined as evidence accumulates. Re-fetch before relying on this; the page above always carries the current score.",
  "scored_at": "2026-08-11",
  "model": "claude-opus-5",
  "occupation": {
    "title": "Textile, Apparel, and Furnishings Workers, All Other",
    "soc_code": "51-6099",
    "category": "Production",
    "us_employment": 13530,
    "median_annual_wage": 37280
  },
  "verdict": "EXPOSED",
  "risk_resistance": 39,
  "contested": false,
  "near_boundary": false,
  "dimensions": {
    "task_resistance": 14,
    "embodiment": 15,
    "liability_shield": 1,
    "trust_premium": 4,
    "judgment_accountability": 5
  },
  "reasoning": {
    "task_resistance": "Aligning nap on a curved seam, easing a sleeve head, re-stitching a hidden defect, hand-tacking linings and pinning a garment on a form all require continuous tactile correction of a material that deforms under its own weight, which is why 14 rather than 18 — the pattern grading, marker making and nesting that used to be part of the bench job has already moved to CAD.",
    "embodiment": "The whole day is standing or leaning at a bench with shears, awl, steam iron, tape and needle on goods that arrive wrinkled, off-grain and dimensionally inconsistent, but it sits at 15 not 19 because the shop floor is a lit, indoor, largely predictable space rather than a job site or someone's roof.",
    "liability_shield": "There is no licence, no state board and no certification that a customer or an OSHA inspector asks to see; a shop can put a new hire on hand finishing the same week, so the 1 reflects only the fact that the work is done by an employed adult at all.",
    "trust_premium": "The finished pillow, drapery panel or garment leaves the plant with a brand label on it and no worker's name, so the 4 comes from the internal relationship — the supervisor or designer who knows which bench hand can be trusted with the expensive fabric — not from any buyer choosing you.",
    "judgment_accountability": "Calls are real but bounded: accept or reject a piece against a spec sheet, decide whether a flaw can be mended or must be scrapped, pick a thread and stitch length for the material — decisions a lead or QC checks the same shift, which is why 5 rather than 9."
  },
  "rationale": "This catch-all bucket covers hand finishing, sample and pattern work, trimming, fitting, mending, and specialty assembly of fabric goods — manual dexterity tasks that language models cannot touch and current robotics still handles badly because limp fabric is hard to grip and predict. The real exposure is not chatbots but the older pressures: purpose-built sewing and cutting automation, digital pattern generation and grading software, and continued offshoring of the work entirely. Scored for the modal worker doing repetitive bench-level fabric handling in a shop or small plant, not the elite sample-room technician.",
  "outlook": "Employment keeps shrinking through the 2030s on offshoring and specialty machine automation, but the workers who survive are those doing one-off fitting, repair, and prototype work that no line can be tooled for.",
  "what_would_raise_it": {
    "levers": [
      {
        "dimension": "task_resistance",
        "change": "Task-mix shift: purpose-built automation (Softwear Automation's Sewbot lines, automated cutting tables) absorbs straight-seam and repetitive bench work first, leaving the residual role as sample-room and prototype fabric handling, one-off repair, and fit correction on non-standard bodies and materials — work with no repeatable geometry for a machine to learn. This tier genuinely exists in the occupation (sample technician vs. bench finisher), so as the routine tier leaves, the surviving job scores higher even with no capability change.",
        "plausibility": "already happening",
        "would_add": 3
      },
      {
        "dimension": "trust_premium",
        "change": "Buyer-side demand for verified hand-made provenance: enforcement of FTC 'Made in USA' labeling rules and growth of certification schemes (e.g. tailoring/bespoke marks, Homegrown/handmade certification used by heritage brands) that let a maker charge specifically because a person did the finishing. Visible now in bespoke tailoring, upholstery restoration, and repair-service growth driven by EU right-to-repair and textile durability rules; if brand repair programs (Patagonia Worn Wear, Eileen Fisher Renew style) expand under EPR mandates, mending becomes a paid human-branded service rather than an invisible cost line.",
        "plausibility": "plausible",
        "would_add": 4
      },
      {
        "dimension": "judgment_accountability",
        "change": "If extended producer responsibility laws for textiles (California SB 707, EU EPR) push repair-and-resale volume into domestic shops, the finisher owns the call on whether a garment or furnishing is salvageable, what substrate and thread to use, and whether the repair meets resale-grade standards — a consequential, ambiguous judgment currently made by no one because the item was simply discarded.",
        "plausibility": "plausible",
        "would_add": 3
      },
      {
        "dimension": "liability_shield",
        "change": "Only narrow route: flammability and safety compliance sign-off on children's sleepwear, mattresses, and upholstered furniture (16 CFR 1615/1616, 1633, California TB 117-2013) where a named person attests to construction conformity. If CPSC or a state fire marshal required an identified finisher/inspector attestation per lot rather than a firm-level certificate, a thin personal-accountability shield appears. No such rule is pending; treat as speculative.",
        "plausibility": "unlikely",
        "would_add": 2
      }
    ],
    "ceiling_note": "The binding constraint is offshoring, not AI. Every lever above can fire and the US headcount still falls if the work moves to Vietnam or Bangladesh — trust-premium and repair-mandate gains apply to a small domestic craft tier, not the modal bench worker, and this SOC bucket is only 13,530 people already."
  },
  "adjudication": null,
  "employment_history": {
    "points": [
      {
        "y": 2017,
        "emp": 14610,
        "wage": 26670
      },
      {
        "y": 2018,
        "emp": 15360,
        "wage": 27750
      },
      {
        "y": 2019,
        "emp": 18060,
        "wage": 27970
      },
      {
        "y": 2020,
        "emp": 16510,
        "wage": 28650
      },
      {
        "y": 2021,
        "emp": 12750,
        "wage": 29750
      },
      {
        "y": 2022,
        "emp": 12810,
        "wage": 32640
      },
      {
        "y": 2023,
        "emp": 15080,
        "wage": 35130
      },
      {
        "y": 2024,
        "emp": 14450,
        "wage": 37010
      },
      {
        "y": 2025,
        "emp": 13530,
        "wage": 37280
      }
    ],
    "from": 2017,
    "to": 2025,
    "change_pct": -7.4,
    "comparable_from": 2019,
    "spans_soc_revision": true
  },
  "pivots": [],
  "license": "https://cookedindex.com/terms"
}