AI Can Do the Work You Hate
Yes, AI can do your laundry. Here’s what that looks like for security teams.
“Routine, repetition, tedium, monotony, ephemeracy, inconsequence, abstraction, disorder, boredom, angst, ennui — these are the true hero’s enemies, and make no mistake, they are fearsome indeed. For they are real.”
– David Foster Wallace, The Pale King
Put this quote in front of any security operations analyst and they will give you a knowing, if beleaguered, look in return. Access Control. Incident reports. CCTV monitoring. Daily metrics. Hours of OSINT monitoring. Then tomorrow, much of the same, all over again. All of it necessary; yet much of it passes without incident, and it is memory-wiped once the shift ends. Unlike other industries, uneventful days require the same amount of attention as the busy days. And after a while, paying attention, real attention, becomes the hardest part of the job. This is the daily reality of a career in physical security.
But Physical Security isn’t unique. Every industry deals with monotony, routine, and tedium. But in our industry, mistakes can be a lot more consequential. We deal in life and limb rather than dollars and cents.
So how do we deal with this? Well, we don’t. We turn gray and go bald working long hours. We burn through junior staff. Tedium is viewed as part and parcel of the job. Paying your dues. And I suspect (but can never know for sure) that the security leaders who make it to the CSO level partly do so because they are excellent at dealing with monotony, and even better at building mistake-proof systems.
But it’s 2026 — surely there’s a different way to work?
The 800 pound artificially intelligent gorilla in the room
A question I sometimes hear AI skeptics ask is: “Yes, but can AI do my laundry?” The answer is, of course, no. Also, ew. But this is less about dirty underwear and more about whether AI can handle everyday, judgement-lite tasks that we’d rather not spend time on, freeing us up to spend more time on stuff we actually want to spend time on.
Physical Security teams have laundry too: minor access control issues, hours of OSINT monitoring, CCTV reviews and collection, routine incident logging, etc. etc. All necessary work, but the repetition makes sustained attention difficult. Anomalies disappear in a sea of ordinary activity, not because people are careless, but because they are tired and bored.
This is where AI comes in. You can be skeptical of AI and still acknowledge that it’s excellent with repeatable, tedious work. It doesn’t get bored, burned out, or call out sick. With proper configuration and oversight, it can power through routine activity, bring exceptions to the human’s attention, and give teams more time to apply their judgement and experience to where it matters.
The future is yesterday
And in some teams, it already is. I was skeptical until I spoke to clients who have deployed AI agents on top of their existing tech stack. Here’s a quick list of what those agents are handling today:
- Dispatching guards to investigate doors-forced/held alarms
- Detecting and escalating incidents captured on video
- Opening maintenance tickets for recurring nuisance alarms
- Automating badge creation and day-one access
- Gathering daily and weekly metrics
- Drafting initial mass communications
If you approached a GSOC lead with the idea that all or even some of this work can be responsibly automated, you’d have a very happy GSOC lead. But how does it actually work?
Take a door-held-open alarm. An analyst might acknowledge the alarm, check the location and camera, dispatch a guard, and document the outcome. Then do it again. And again. An agent can be configured to gather the relevant information and follow the team’s established procedure. For a routine alarm, that might mean dispatching a guard. For an alarm at a sensitive location, it might mean flagging it for an analyst to review.
Of course, this requires a bit more than waving a magic AI wand over the GSOC. Teams need to decide which steps the agent can handle, when it should involve a human, and how they’ll check its work. The same goes for that draft mass communication: the agent can put the first version together, but someone with the right authority should review and send it.
AI can’t replace experience and judgment. But it can do your team’s proverbial laundry, giving analysts more time and attention for the things that need both.