I’ve walked into more galleries after a break-in than I care to count, back when I was running operations for a mid-size Melbourne gallery. The thing that always surprised the owner was never the thing they’d been worried about. They’d spent money on an alarm. They’d spent money on cameras pointed at the front door. And the loss almost always traced back to somewhere else entirely, usually somewhere nobody had thought to look.
It’s Not the Overnight Burglar You Should Be Worried About #
Ask most gallery owners what keeps them up at night and they’ll describe the same scene: someone smashing a window at 2am, grabbing a painting, gone before the monitored alarm even finishes dialling out. It happens, and it’s worth planning for. But in my experience it’s a smaller slice of the real risk than people assume, and chasing it hard while ignoring everything else is how a lot of galleries end up exposed in the exact spot they never checked.
The bigger, quieter problem is access: who can actually get near the stock, during hours when nobody’s paying close attention, using a key or a code that was handed out months ago and never tracked. It doesn’t make for a dramatic story, so it rarely gets the attention it deserves.
The Frans van Mieris Painting That Walked Out in Daylight #
The clearest example I know of isn’t even a break-in. In June 2007, a small 17th-century panel painting by Frans van Mieris, “A Cavalier”, was stolen from the Art Gallery of New South Wales in the middle of the day, while the gallery was open to the public. No forced entry. No alarm to defeat. Someone simply took it off the wall and walked out, and it’s never been recovered. It’s still listed on the FBI’s Top Ten Art Crimes, nearly two decades on.
I bring this up constantly with clients because it’s the case that actually matches how most losses happen. Not a dramatic overnight raid, but a quiet failure of who’s watching what, during hours everyone assumed were safe. A gallery open to the public is, by definition, a gallery where a stranger can walk right up to a piece and nobody finds that unusual until it’s too late.
Where the Real Gaps Sit: Keys, Not Locks #
Here’s the mildly unpopular opinion, and I’ll stand by it after eleven years of doing this: cameras get too much of the security budget and keys get almost none. Every gallery I’ve worked in has had a drawer, or a hook, or a lockbox with a set of keys that’s been copied at some point by someone nobody can quite remember, for a reason nobody can quite remember either. Contractors. Former staff. A framer who did one job back in 2019 and was handed a key “just for the day” that never came back.
A lock is only as good as the record of who has a key to it, and in most galleries and dealerships that record doesn’t exist. It’s not written down anywhere, its just assumed to be fine, right up until something goes missing and there’s no way to even narrow down who could have taken it.
What Proper Key Control Actually Looks Like #
Electronic access control fixes this in a way a physical key never can, because it logs the event. Not “we think Dave still has a key” but an actual timestamped record of who opened which door, and when. For a working gallery that’s genuinely useful in a way a padlock isn’t. If a piece goes missing from storage, you’re not left guessing at a list of everyone who’s ever worked there.
The layering matters too. Staff, contractors and visiting valuers don’t need the same access, and most galleries I’ve audited had given everyone the same key regardless of role, because setting up different levels felt like more admin than it was worth. It isn’t, once it’s actually in place, and the difference shows the first time something needs investigating rather than guessing.
The Standard Everyone Quotes and Nobody Quite Follows #
The National Standards for Australian Museums and Galleries, now in its second edition, sets the closest thing this sector has to an official benchmark. Its actual wording is that an organisation “meets fire safety standards and uses appropriate security measures.” That’s it. No CCTV spec, no alarm-response time, no mention of key control at all. It’s a soft standard by design, which means every gallery gets to decide for itself what “appropriate” means, and in my experience that usually means whatever was already in place when the standard was written, rather than anything genuinely reassessed.
I don’t think that’s a criticism of the standard so much as a warning about how it gets used. A soft benchmark is easy to technically meet while still leaving a real gap sitting in plain sight. I’ve sat in meetings where a director pointed at the alarm contract and the camera install as evidence the box was ticked, and nobody in the room could tell me how many sets of keys existed for the storeroom two doors down. Both things were true at once: technically compliant, and genuinely exposed.
That’s the pattern I keep coming back to. A standard that doesn’t specify hardware ends up measured by whatever’s easiest to point to, and keys are never easy to point to, because there’s rarely a single document listing them. CCTV footage is visible. An alarm contract is a piece of paper you can wave at an insurer. A set of keys handed to a contractor in 2019 is not, and that asymmetry is exactly why it gets skipped.
After Hours Is Still Real, Just Not the Whole Story #
None of this means overnight risk doesn’t matter. It does, and a properly zoned monitored alarm covering storage and display separately from the public floor is worth having, especially once the doors are locked and nobody’s around to notice something’s wrong in real time. I’m not arguing against alarms. I’m arguing against treating them as the whole answer, which is what happens when a soft standard meets a limited budget and everyone reaches for the obvious fix first.
For dealers running a smaller operation without a dedicated storeroom, the same logic applies at a smaller scale. Even a single back office holding consigned pieces benefits from the same question: who actually has a key, and would you know if that changed?
What I’d Actually Check First #
If I were walking into a gallery today to do a proper assessment, which is more or less my job now, I wouldn’t start with the alarm panel. I’d start by asking who has keys, how many sets exist, and when the locks were last changed. Nine times out of ten nobody has a confident answer, and that’s the actual gap, sitting there quietly while everyone’s attention is on the front door.
It’s not a glamorous fix. It’s also usually cheaper than the camera upgrade everyone reaches for first, and it closes a hole a camera was never going to close in the first place. If you run a gallery or dealership and genuinely can’t answer who holds a key to your storeroom right now, that’s worth sorting before anything else on the list.