{
  "source": "Cooked Index — occupational AI risk register",
  "page": "https://cookedindex.com/jobs/maids-and-housekeeping-cleaners/",
  "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": "Maids and Housekeeping Cleaners",
    "soc_code": "37-2012",
    "category": "Facilities",
    "us_employment": 860670,
    "median_annual_wage": 35510
  },
  "verdict": "EXPOSED",
  "risk_resistance": 54,
  "contested": false,
  "near_boundary": false,
  "dimensions": {
    "task_resistance": 18,
    "embodiment": 19,
    "liability_shield": 1,
    "trust_premium": 11,
    "judgment_accountability": 5
  },
  "reasoning": {
    "task_resistance": "An 18 reflects that essentially every duty — pulling soiled linens off a mattress, scrubbing tub grout, moving nightstands to vacuum behind them, wiping baseboards, hauling trash to a chute — has no text or data analogue at all, and the handful of digitizable pieces (marking a room clean in the PMS tablet, logging a lost-and-found item) take minutes out of an eight-hour shift.",
    "embodiment": "A 19 is warranted because the work happens in rooms no two of which are laid out the same after a guest has used them: unpredictable spills, furniture shoved around, cords, pets and children in private homes, wet tile, and constant kneeling, reaching over beds, and pushing a loaded cart down uneven corridors — conditions no fixed robotic footprint or mapped environment covers.",
    "liability_shield": "A 1 is right because nothing about entering a stranger's bedroom and handling their possessions requires a state credential; the only formal gates are OSHA hazard-communication training on chemical labels and a background check the employer runs, and if a guest's property goes missing the hotel's insurer and management absorb the claim, not you.",
    "trust_premium": "An 11 sits above hotel-floor anonymity because a recurring private-home or long-stay client knows your name, gives you a key or code, tells you which drawers not to open, and will not accept a substitute cleaner without complaint — but it stays below 13 because in large hotels the buyer books a room, not a housekeeper, and floor assignments rotate you across guests who never learn who cleaned.",
    "judgment_accountability": "A 5 matches a job run on written room-cleaning sequences, brand standards for towel folds and amenity placement, and a supervisor's inspection checklist; the discretion you do hold — flagging a bed bug, deciding a stain needs the carpet team, choosing not to disturb a DND room — is real but escalates upward rather than being your call to close."
  },
  "rationale": "Almost nothing in this job is text or screen work: stripping beds, scrubbing bathrooms, vacuuming around furniture, restocking carts, and handling guest requests are physical tasks in cluttered, variable spaces that current robotics cannot do end-to-end. The real exposure is not the AI doing the cleaning but AI-driven scheduling, room-assignment optimization, and demand forecasting squeezing headcount per room and shortening allotted minutes per unit. There is no license and little discretion, so the protection here is purely physical, and private-home cleaners hold more of a trust relationship than hotel-floor staff, who are largely interchangeable to the buyer.",
  "outlook": "The work itself stays human through the 2030s, but expect fewer cleaners per property as software tightens room quotas and minutes-per-unit, with the best pay shifting to private-client and specialty-cleaning niches.",
  "what_would_raise_it": {
    "levers": [
      {
        "dimension": "trust_premium",
        "change": "Expansion of state hotel-housekeeping workload laws (e.g., NV SB 4-style panic-button and room-quota rules, Seattle I-124, LA's Hotel Worker Protection Ordinance capping square footage per shift and mandating overtime pay above it) that make the buyer pay explicitly for a named, assigned human per floor rather than a pooled algorithmic dispatch; plus continued growth of vetted-named-cleaner platforms in private homes where repeat-client matching is the product",
        "plausibility": "already happening",
        "would_add": 3
      },
      {
        "dimension": "judgment_accountability",
        "change": "Formalizing housekeepers as the reporting node for guest-safety events — human-trafficking indicator training mandated for lodging staff (already law in CA AB 2034/ECPAT-style statutes and several state hotel licensing rules), infection-control sign-off after isolation-room turnover in healthcare/assisted-living cleaning, and documented chain-of-custody for lost-and-found and controlled items. If the room turnover record becomes an attestable document the cleaner signs, the role owns a consequential call",
        "plausibility": "plausible",
        "would_add": 4
      },
      {
        "dimension": "liability_shield",
        "change": "Certification-gated cleaning niches becoming mandatory rather than optional: EPA RRP lead-safe certification for pre-1978 dwellings already requires a certified individual, and state-level requirements for terminal-clean or biohazard/crime-scene remediation credentials (e.g., trauma-scene practitioner licensing in CA and FL) put a named certified person on the record. Broadening a healthcare-environmental-services credential requirement (AHE CHEST-style) into CMS conditions of participation would attach personal accountability to hospital cleaning",
        "plausibility": "plausible",
        "would_add": 4
      },
      {
        "dimension": "task_resistance",
        "change": "Task-mix shift is limited but real where robotic floor scrubbers absorb the open-floor vacuuming tier (Tennant/Brain, Avidbots deployments in hotels and airports), leaving the residual job concentrated in edge-cleaning, bathrooms, linen handling and guest interaction — the tier robots handle worst. This raises resistance of what remains while reducing headcount",
        "plausibility": "already happening",
        "would_add": 1
      }
    ],
    "ceiling_note": "Even with every lever, the binding constraint is headcount, not capability: AI squeezes minutes-per-room and staffing ratios without touching the physical task, so the score can rise while employment falls. Trust premium is structurally capped in hotels, where the buyer never meets the cleaner; the private-home segment is where it lives, and it is the smaller share of these 860k jobs."
  },
  "adjudication": null,
  "employment_history": {
    "points": [
      {
        "y": 2017,
        "emp": 922660,
        "wage": 22860
      },
      {
        "y": 2018,
        "emp": 924290,
        "wage": 23770
      },
      {
        "y": 2019,
        "emp": 926960,
        "wage": 24850
      },
      {
        "y": 2020,
        "emp": 795590,
        "wage": 26220
      },
      {
        "y": 2021,
        "emp": 723430,
        "wage": 28780
      },
      {
        "y": 2022,
        "emp": 771390,
        "wage": 29960
      },
      {
        "y": 2023,
        "emp": 836230,
        "wage": 33450
      },
      {
        "y": 2024,
        "emp": 854910,
        "wage": 34660
      },
      {
        "y": 2025,
        "emp": 860670,
        "wage": 35510
      }
    ],
    "from": 2017,
    "to": 2025,
    "change_pct": -6.7,
    "comparable_from": 2019,
    "spans_soc_revision": true
  },
  "pivots": [],
  "license": "https://cookedindex.com/terms"
}