Maintenance photo documentation that holds up in a dispute

California just started requiring it

In 2024 California passed a law that shows where this is going. Under AB 2801, a landlord who takes money out of a security deposit now has to photograph the unit — before move-in, after move-out, and after any repair the deduction pays for — and send those photos with the itemized statement. The statute it amended, Civil Code section 1950.5, already gives the tenant an itemized statement within 21 days and the right to challenge it in small claims court, with damages up to twice the deposit for a bad-faith violation.

One state, one kind of deduction. But the direction is one way. A photo is turning into the thing you are asked to produce, not a nicety you keep if you remember. So it is worth getting right, because a photo that cannot be trusted is not much better than no photo at all.

A phone photo is not a record

Here is the problem with how most maintenance photos get taken. A worker snaps a picture and texts it to the office. It shows something got done. It does not show when, and it does not show where. The date on a photo file is set by the phone, and a phone’s date can be changed in seconds. Send the photo through a messaging app and the app recompresses it and strips most of the little context it had. Now it lives in a chat thread on someone’s personal phone, mixed in with everything else.

Months later a tenant says the repair never happened. You go looking for the photo. Maybe you find it. Maybe the worker got a new phone. Either way, the photo proves a picture exists. It does not prove the work did. That is the gap, and it always shows up at the worst time.

What makes a maintenance photo hold up

A photo holds up when it stops being a loose file and becomes a record. That takes four plain things, captured at the work, not typed in afterward:

• A live capture. The worker taps a link and the camera opens; gallery uploads are blocked, so the photo is taken at the unit, not pulled from a camera roll.
• A location. GPS is recorded at the moment of the photo and tied to the site.
• A server-set time. The timestamp is written by the server, not the handset, so the worker cannot set it.
• A chain of custody. A SHA-256 fingerprint seals the record, so any later change to it is detectable.

Put those together and file each one by unit and date, and a maintenance photo becomes evidence-grade instead of a picture you hope someone kept. Here is how the common ways of keeping photos compare on the points that decide a dispute.

Where it earns its keep

The value of documentation shows up at four moments, and they are the expensive ones. A tenant files a habitability complaint saying the heat was never fixed; with a dated, located photo of the finished repair, the complaint meets a record instead of your word. A deposit deduction gets challenged at move-out; the before-and-after condition is already filed, which is exactly what AB 2801 now asks for. You file a property-damage claim, and the adjuster wants dated proof of the condition and the repair — a record with a sealed time and location is cleaner to hand over than a folder of undated shots. And an owner asks whether the work they paid for got done; you pull the job and share it.

Say an owner questions a $1,200 invoice for a unit turn across town — a hypothetical, but a familiar one. You open the job, and there it is: the finished unit, the location it was captured, and the time, filed under that address. The conversation is over in a minute instead of a week. That is the whole point of proof of repairs for a tenant dispute or an owner question: you stop arguing and start showing.

Keep the photo, own the record

A maintenance photo is only worth what it can prove. Take it at the work, tie it to a place and a sealed time, and keep it where you can find it. Then a disputed repair is a record you pull, not an argument you have. See what one finished record looks like in a sample verified record.