Why the water and the mountain still need eyes and hands
Watching is only the first half of this job. A guard scans a crowded pool or a surf line, spots a swimmer in trouble, then enters the water, brings the person out, and starts first aid. A ski patroller inspects trails, closes hazards, runs avalanche control work, and gets an injured skier down the hill in a toboggan. Software can flag a problem. It cannot carry anyone.
That is why the share of task time that still needs a person here is 92%. The rescue itself is physical, wet, cold, and improvised. So is splinting a broken leg on a slope in bad light, or clearing a lightning-struck pool deck while parents argue. These are the tasks that set the score, not the paperwork around them.
There is a second reason the work stays with people: someone has to be accountable on site. Pools, beaches and resorts run under staffing rules, certifications and insurance terms written around trained humans. A camera that misses a swimmer is a product failure. A guard who misses one is a person with a license, a supervisor and a report to file.
What AI does, what it helps with, and what stays with the guard
The slice of task time AI already handles on its own is 0%. It sits at the desk end of the job: writing up incident reports from structured details, and keeping attendance, patrol and equipment logs that used to be filled in by hand at the end of a shift. That work matters for liability, but it is a small part of a day. You can see how we measure that share on the coverage method page.
The assisted share is 8%. This is where most of the new technology lands. Underwater and overhead camera systems watch for a motionless body and sound an alert, which supports the task of monitoring swimmers rather than replacing it. On the mountain, modeling tools support avalanche hazard assessment by pulling weather, wind loading and snowpack data into one forecast. Drones can reach a struggling swimmer with a flotation device faster than a guard can swim.
Everything else stays with people. Pulling a distressed swimmer out and giving CPR is one of those tasks. So is enforcing rules with patrons who do not want to hear them, checking and maintaining rescue equipment, and deciding when conditions mean the beach closes or the run opens. Judgment under noise and time pressure is the job, and no alert system takes it over.
How strong is the evidence?
Weaker than people assume. Our evidence grade for this occupation is D, which means no published study has tested an AI system against a qualified lifeguard or patroller on the core tasks. Detection vendors publish their own claims; those are not independent tests, and we do not score them. Because the grade is unmeasured, we give no parity number at all. The quality parity method explains why a number without a test is worse than no number.
What would settle it: field trials at real pools and beaches that count missed events and false alarms for a detection system and for trained guards over the same shifts, plus response-time and outcome data after an alert. Until something like that is published and dated, the honest position is that the technology supports surveillance and has not been shown to match a person at it.
When the balance could change
Most likely after 2042 (8 in 10 of our scenarios). Our overall coverage figure for the occupation is 8 out of 100, and the way that window is built is set out on the replacement year method page.
Two things could pull the date earlier. The first is cost: monitoring hardware and software are cheap next to a staffed shift, so operators facing hiring shortages have a reason to buy them. The second is drones and remote flotation delivery, which already shorten the gap between spotting a swimmer and reaching one.
Two things hold it back. The physical share of this job needs a dexterous humanoid machine to touch at all, and nothing at that tier works in surf, cold water or deep snow. And the rules do not bend quickly: staffing ratios, certification and insurance all assume trained people on deck. Our wider look at humanoid robots and physical jobs covers why that tier stays slow.
How to stay needed in recreational safety work
Lean into the tasks the machines do not touch. Keep your water rescue and toboggan handling sharp. Keep first aid, CPR and oxygen certifications current, and add the higher ones your site uses, such as EMT or outdoor emergency care. And take on the patron-facing side: rule enforcement, hazard briefings and incident debriefs are judgment work that supervisors notice.
Two skills raise your value fast. One is incident command and training others, because the person who runs drills and writes the response plan is harder to cut than the person who only sits the chair. The other is working with the alert systems themselves: knowing their blind spots, calibrating them, and deciding when to trust them is a supervisory skill in itself.
What to do: ask your employer whether any detection or drone system on site has been tested against staffed coverage, and get yourself on the team that reviews it.
If you are weighing your next step, nearby work is worth a look: Security Guards, Firefighters and Recreation Workers all share parts of this job’s mix. You can put any two side by side with the job comparison tool, see the rest of the other protective service workers family, or look at the arts and entertainment sector page for the employers that hire most of these roles. Our list of jobs that mostly need a person shows where this kind of work sits against everything else, and the full scoring method shows how each figure above is built.
For context on pay and demand: the BLS counts about 157,550 US workers in this occupation and a median wage of $33,580, with employment projected to grow 5.5% from 2025 to 2035 (BLS, 2025). That is a job market shaped by seasons and public budgets far more than by software.