If you're like me, you've read The Art of War. Or at least you've sat through a keynote where someone put it on a slide. I'm a book geek, so I've done both. Sun Tzu has been applied to sales, chess, football, and probably your last reorg.
Here's the thing. Most of it reads like common sense. Know your enemy. Know yourself. Don't pick fights you can't win. Thanks, coach.
I recently came across some analysis that changed how I read it. When it was written, that kind of common sense was anything but common.
A quick note for the peanut gallery. Sun Tzu probably didn't write The Art of War. Not alone, anyway. Tradition credits a general from around 500 BC, but most scholars think the book was compiled later, during the Warring States period, likely by several hands. Some doubt Sun Tzu existed at all. I'll keep calling it his book. Everyone else does. And the later date actually helps my case. It was written by people living through the very shift it describes.
War as a sport
Some background. War in Sun Tzu's China looked a lot like medieval Europe. It belonged to the nobility. Rich landholders with the advanced technology of the day, the chariot, rode out to settle vendettas, answer insults, and win honor. There were rules. Etiquette. You fought the right way or you didn't fight at all.
My favorite example is the Duke of Song. In 638 BC his army faced a larger force from the state of Chu, which had to cross a river to reach him. The Duke’s officers begged him to attack while the enemy was mid-crossing. He refused. Not sporting. They begged him to attack before the enemy formed ranks. Refused again. A gentleman doesn't strike an unready opponent. Chu formed up, routed his army, and the Duke took an arrow in the thigh. He died of the wound the next year.
Honor intact, though.
Read Sun Tzu against that backdrop and he stops sounding like a fortune cookie. He sounds like a prosecutor. War should serve the state, not a noble's ego. A ruler should never go to war out of anger, and a general should never fight out of spite. No country has ever benefited from a long war. And supreme excellence is breaking the enemy's resistance without fighting at all.
I can just picture the nobles reading that. Victory without bloodshed? What's the point? The bloodshed is the best part!
Sun Tzu was calling time on war as a sport for aristocrats. The era that produced the book replaced it with something new. Mass armies. Professional generals. Logistics, intelligence, deception. War as a serious, state-wide enterprise. The chariot crowd didn't make the cut.
The chariots of cybersecurity
What does this have to do with AI?
We're living through the same kind of shift in security. For twenty years most defense programs have been built around two goals: pass the audit, and don't disrupt the business. Quarterly scans. Annual pen tests. Patch Tuesday, then a change window, then a meeting about the change window. The policy binder is our chariot. Expensive. Impressive. Built for a war that's ending.
Attackers are already using AI to find weaknesses, write exploits, and run campaigns faster than any human team. The gap between a flaw being disclosed and being exploited keeps shrinking. If your defense moves at the speed of a quarterly review, you've shown up, in full armor, for a fight that ended last Tuesday.
Sun Tzu had a line for this too. Rapidity is the essence of war.
Look, I'm not saying compliance doesn't matter. It does. The Duke of Song had rules too. Rules aren't the problem. Mistaking the rules for the war is the problem.

Fighting the new way
So, what does the professional army look like? Here's my attempt at a modern edition.
Know what you're worth. Every attack is a business decision. Somebody on the other side is doing the math: what will it cost to get in, and what's it worth once I'm there? You should be doing the same math. Most companies never have. Ask a board what its data is worth to a criminal and you'll get a long pause.
Know what it costs to attack you. That's the other half of the equation. Sun Tzu said know the enemy and know yourself. Our version is know what you're worth to them, and know what you cost them.
Make your opponent spend more than you're worth. That's victory without bloodshed, updated. You don't need to be unbreakable. You need to be a bad investment. The attacker probes, the math doesn't work, and they go bother someone else. Nobody issues a press release about the breach that didn't happen. Best outcome anyway.
Prepare for attacks that cost almost nothing and never stop. This is where AI changes the math. When an attack took a skilled team weeks, attackers picked their targets carefully. When an AI agent can probe you for pennies—around the clock—everyone is a target and the probing never ends. You're not defending against a siege anymore. You're defending against weather. Defense that waits for a human to open a ticket loses. AI has to be working on your side of the line too, triaging, containing, adapting. People still make the calls that matter. They just stop being the bottleneck for the calls that don't.
We can protect anything. We can't protect everything. That's not a budget problem. It's a platform decision. Pick a cloud, or two, and a way of running in it, whether that's lift and shift, containers, or serverless, and we can secure it well. Try to run every cloud with every strategy at once and nobody can. AI is the same. We can protect the models you've chosen, the data you've decided to put behind them, and the people you've given access on purpose. We can't protect every model, every data set, and every user with free rein to play. That's not a strategy. That's the play-around-and-find-out school of IT.
And the sprawl doesn't wait for permission. New SaaS apps. New identities. New AI agents with access nobody remembers approving. An annual asset inventory is a portrait of a house that's already been remodeled. So the attack surface has to be managed as a business decision, with one cohesive strategy. Choose your platforms. Decide what matters most. Spend where the math says to spend. Shrink exposure everywhere else, continuously. That's not a security project. That's how the business runs now.
The skeptic in the room will say we've heard this before. Cloud was going to change everything. Zero trust was going to change everything. Fair. But the Duke of Song had heard about the new tactics too. He just thought they were rude.
The short version
The real lesson of The Art of War isn't any single maxim. It's that the nature of war changed, and Sun Tzu noticed before the people holding the reins did. The nobles who couldn't adjust didn't lose an argument. They lost their states.
Twenty-five hundred years later, same lesson. Different chariots.
Researched and drafted with Claude, which should surprise exactly no one given who I am. The arguing and editing are mine.