Will AI replace other service workers not elsewhere classified?
Generative AI exposureMinimal exposure
Overall exposure is limited, although some tasks have moderate potential for AI.
Average task score: 0.26 / 1 · ILO 2025
Tasks and how exposed they are
Each task is scored from 0 (current generative AI cannot do it) to 1 (AI could do it entirely). Scores show technical potential, not whether the task will actually be automated.
- Reading electricity, gas or water meters and recording consumption0.75 · High
- Planning and following the most efficient route0.70 · Medium
- Keeping records of merchandise distributed and money collected0.58 · Medium
- Selling admission tickets and passes to patrons at entertainment events or collecting tickets, passes and tags from patrons0.48 · Low
- Examining tickets or passes to verify authenticity, using criteria such as colour and date issued0.45 · Low
- Sorting items to be delivered according to the delivery route.0.42 · Low
- Receiving and marking baggage by completing and attaching claim checks0.41 · Low
- Calculating parking charges and collecting fees from customers0.38 · Low
- Verifying readings in cases where consumption appears to be abnormal, and record possible reasons for fluctuations0.35 · Low
- Receiving immediate payment.0.34 · Low
- Directing patrons to restrooms, concession stands and telephones0.33 · Low
- Proceeding along established routes to take readings of meter dials0.32 · Low
- Inspecting meters for unauthorized connections, defects and damage such as broken seals.0.27 · Low
- Assisting car drivers to find a parking place and ensuring that the car is not damaged during driver’s absence0.26 · Low
- Delivering various goods to and from enterprises, shops, households and other places0.25 · Low
- Delivering messages, packages and other items within an establishment or between establishments, or elsewhere0.24 · Very low
- Directing vehicle drivers to parking spaces0.20 · Very low
- Carrying and delivering luggage at hotels, stations, airports, and elsewhere0.20 · Very low
- Obtaining the materials necessary to perform services0.20 · Very low
- Collecting money from parking meters and similar coinboxes0.19 · Very low
- Patrolling parking areas in order to prevent vehicle damage and vehicle property thefts0.19 · Very low
- Running errands0.18 · Very low
- Filling storage areas of vending machines and collecting money from their containers0.18 · Very low
- Assigning dressing room facilities, locker space or clothing containers to patrons of athletic or bathing establishments.0.17 · Very low
- Approaching people on the street to offer services0.15 · Very low
- Adjusting doors and windows0.15 · Very low
- Replacing defective items such as light bulbs0.14 · Very low
- Cleaning and polishing shoes0.14 · Very low
- Guiding patrons to exits or providing other instructions or assistance in case of emergency0.14 · Very low
- Repairing and painting interior and exterior surfaces such as walls, ceilings and fences0.12 · Very low
- Unloading coal or wood and putting it into cellars of private households or establishments.0.10 · Very low
- Repairing broken windows, screens, doors, fences, barbecues, picnic tables, shelves, cupboards and other items0.10 · Very low
- Cleaning and polishing car windows0.08 · Very low
- Handing out leaflets and free newspapers0.08 · Very low
- Putting up handrails and grab rails0.08 · Very low
- Replacing tap washers0.08 · Very low
How people actually use AI for this work
Share of tasks already seen in real AI usage: 1%
21%
79%
Augments (helps the worker) Automates (does the task)
Source: Anthropic Economic Index
Job facts (USA)
- Employed
- 186,750
- BLS 2025
- Median annual wage
- Coming soon
Similar exposure
Data updated: 2026-11-18 · SSYK 9629 · ISCO-08 9510, 9621, 9622, 9623, 9629 ·Methodology