> ## Content Index
> Fetch the complete content index at: https://www.nanosava.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Proof of Lie
- URL: https://www.nanosava.com/books/proof-of-lie/proof-of-lie-essay/
- Published: 2026-10-13T04:17:04.000Z
- Updated: 2026-10-15T07:25:01.000Z
- Description: When thinking gets cheap, the last advantage in cybersecurity is not a smarter machine. It is knowing what it would cost to fool one.
- Author: Sava
- Tags: Proof of Lie, Essay

Proof of Lie · Prologue

In August 2025, two rooms at DEF CON told the same story from opposite ends. In one, DARPA's AI Cyber Challenge put autonomous systems on stage to read unfamiliar code, find the flaws, and prove the patches held, with no human at the keyboard. A few halls away, in the AI Village's orbit, machines ran the other direction: mapping a target, chaining its weaknesses, breaking in, and getting sharper with every failure. Two rooms, one fact. Both sides now learn at machine speed.

That month quietly retired the way the security industry had organized itself for a generation. For thirty years we bought controls and counted them: antivirus, firewalls, the SIEM, the EDR, a product for every verb in detect, prevent, respond. Underneath all of it sat an assumption so old nobody said it aloud, that the defender was the bigger, better-resourced mind and the attacker could not out-think a well-run shop. When expert reasoning costs almost nothing and is available to everyone, that assumption is gone. The bigger brain stopped being something you could own.

## The moat moved from tools to loops

For those with booths at that conference, the uncomfortable truth is that every previous defensive product was an artifact, and every artifact teaches. A detection signature encodes what an analyst learned about a family of attacks; publish it, or let it fire once, and it tells the adversary exactly what to evade. A tuned classifier is a compression of yesterday's labels. A patch is the fossil of a fixed bug. Each is real work. None of it is a moat, because each can be reconstructed, routed around, or simply out-iterated by the same cheap cognition the other side just bought.

What does not commoditize is the loop that produces the next capability faster than the adversary retires the last one: capture what happens, connect each decision to how it turned out, ground the models in what has actually been validated, and improve. A tool has a coverage number. A loop has a rate and a direction. Those two properties are the only ones cheap cognition does not level, because they come from operating a real environment over real time, and that is the one input a competitor cannot download.

So this new contest is a race between two learning systems, and this race between learning systems has an ugly symmetry: if both sides think at machine speed on the same commoditized models, how does either one ever win?

## Measuring is the new question of cybersecurity

Decades ago, cryptography effectively used cost or effort to measure resilience. A modern cipher does not ask you to trust that its designers were clever. It tells you, in bits, what breaking it must cost. 

*"This approach uses 4096 bits and will take 20 billion years to break."*

The promise is not "we tried hard." The promise is a work factor, stated in advance and generally independent of how ingenious the attacker is. 

Cybersecurity has never had that number for the thing that now matters most: not whether your machine was compromised, but whether its judgment was. In today's world of agents, can you ask trick them to believe something differently? Can you *lie* to it and change its mission or open up a gap? Now instead of the number of bits in a cipher, you ask, "what does it cost to make a learning machine believe a lie, and keep believing it?"

Today's answer is already humbling for a previous generation. Recent research, still lab work rather than fielded practice and carried with that caveat, has found two regimes. Poison a retrieval corpus and a learned decision within an agent system has flipped at about one document in two thousand. Poison pretraining, and about 250 documents have planted a hidden trigger, a count that did not grow with the model across the sizes tested. Those experiments measure only the cost of a flip; they never ask the lie to survive re-checking, and neither do most deployed systems. Count the cost in independent sources rather than raw documents, and the honest figure is a ceiling: in cryptography's vocabulary, the work factor of a lie is at most about one. One cheap, well-placed input can buy a consequential decision, and deleting it afterward undoes none of what it taught.

We would never ship a cipher that breaks on the first try. We would not call it a cipher. Yet that is roughly the integrity guarantee under which the industry is now wiring autonomous systems into its defenses. The scandal is not that poisoning works. It is the price.

## Make the lie prove itself

Some may notice that the title of my upcoming book *Proof of Lie* is a play on *proof-of-work*, the mechanism that makes a blockchain claim expensive to assert. An agentic learning loop built correctly should do the same thing to deception. Before the machine believes, a falsehood should have to supply its proof of lie: the volume of independent-looking reality an adversary must forge to push a conclusion over the loop's threshold. The whole program of a serious defense, in one sentence, is to raise that price from one toward something a liar cannot afford, and to make everything under it reversible.

And here is the turn that makes the fight winnable, because it does not depend on having a smarter machine than the other side. It depends on a property of information itself.

Truth is redundant. A lie is not.

A true state of the world is corroborated from many uncorrelated directions, because reality leaves evidence everywhere it touches. Accumulating that evidence is cheap for whoever is telling the truth. A lie has no such backing. To be believed, it must manufacture the independence it does not have, forging enough uncorrelated-looking sources to cross the line, and then hold that forgery in place against a loop that keeps re-checking its conclusions and revoking whatever fails to corroborate. Corroboration is cheap for truth and expensive for a lie, and the gap between those two costs is the defender's edge. It is not a feature you can buy off a shelf. The redundancy it exploits is a property of information; the edge itself has to be built, and it holds only as long as your loop is engineered to tell independent sources from manufactured ones.

## Truth and Lies: A new era of learning Cybersecurity

When intentional, manufactured lies spread to an ecosystem, we enter an era where protecting intent and "truth" in our agentic systems becomes the critical defense. In 2026, reporting based on leaked planning files of a Kremlin-contracted agency described plans for Wikipedia-style reference sites, media outlets, and phony think tanks, built to capture search traffic and steer what AI chatbots repeat as fact. It is recognition that when cognition is cheap, the high ground is no longer any model, it is the corroborated record that everyone's models are trained and grounded on, and it is under deliberate attack.

Closer to home, there is the problem every metric in security eventually hits: the moment a number (like the # of bits to keep a cipher safe) decides budgets, it becomes a target for everyone involved. Three sets of hands generally reach for it: the defender's, the vendor's, and the adversary's. This surfaces the hardest question for the industry: Who grades the defenders when almost every grader is paid by the graded? The autonomous pentester certifies its own coverage. The AI SOC grades its own triage. The benchmark is run by the party it ranks. Independence, it turns out, is the one thing no vendor can supply for itself, and the scarce asset the whole industry is short of.

## What comes after the smart machine

None of this is a counsel of despair, as the dynamics are evolving on all side of the cybersecurity tension. Every advantage that cheap cognition erases for the defender, it erases for the attacker too, and what survives this leveling will structurally favor the side that lives closest to the truth. Reality keeps handing out its corroboration almost for free. For a loop built to price independence, checking a claim stays cheaper than forging one. The contest after intelligence is not who thinks best. It is who can prove what they know, price what a lie would cost, and afford to unlearn what turns out to be poison.

The new moat was never the smarter machine. It is knowing what it costs to fool one, and building so that a lie stops being cheap.

---

*Sava is the author of* Proof of Lie: Learning Faster Than the Adversary in AI-Speed Cybersecurity *(Judgement Press, publishes December 1, 2026): how a modern defense learns, where that learning can be poisoned, how to measure both, and who has the independence to say any of it is real.*