The Harm We All Sense But Couldn’t Name — Until Now

I’m the founder of an AI company. Not a critic on the outside looking in. That’s the context for everything that follows.

The mechanisms of manipulation are ancient. But technology has given those with money, power, or the skills something new — reach at scale and individualized targeting at the same time. It’s been weaponized and used against us for profit and power. And right now, there is no law anywhere on earth that names this as a crime.

Not fraud law. Fraud requires a discrete victim, a specific transaction. This operates at population scale — millions of people, no single identifiable victim.

Not defamation. Defamation requires false statements about specific people. This manufactures a false picture of the world itself.

Not incitement. Incitement needs a specific trigger moment. This is ambient. Cumulative. By the time it’s visible, it’s already done its work.

The legal architecture of every functioning democracy was built before mass social media, before AI-generated synthetic content, before coordinated inauthentic behavior became a purchasable service. The gap isn’t an oversight. Nobody imagined what systematic epistemic coercion at industrial scale even looked like.

Now we know what it looks like. And the trajectory is clear.

A society without shared objective truth stops disagreeing and starts dehumanizing. Those who disagree become the enemy. And when they are no longer heard or seen as human, they are to be hated, feared and loathed. History is very unambiguous about where this leads — and it’s a very ugly place that I hope the children of our planet never have to see or experience again.

We didn’t choose this. It was engineered. And it’s increasing.

I’ve been designing and building systems for over 30 years. I know how feedback loops work.

People talk about “blast radius” — in politics, in business, any time something or a situation blows up. The term is military. Nuclear weapons made it part of the cultural vocabulary — the idea that you could map the damage. Get clear of it.

The people who should be talking about the blast radius of getting AI wrong haven’t been.

It’s all of humanity. There is no outside. No geography puts you clear of it.

So what would a law that actually covered this look like?

Manufactured Psychological Coercion is the production or dissemination of information, in any form, that corrupts the information environment for benefit — perceived or real — where prior knowledge of its falsity existed or a sufficient duty of care should have been exercised. The information must be demonstrably false, a deliberate distortion, or any other misrepresentation of objective truth.

The larger the platform, the higher the duty of care. The reach that comes with the position is not a shield — it is a responsibility.

Personal benefit from the false narrative is admissible as evidence. The conduct continuing unchanged after confrontation with evidence is not a defense — it is further evidence.

Those who repeat a false narrative in genuine belief are not perpetrators — they are evidence of its reach. That changes the moment the individual knows the originator has recanted or evidence has been presented to the contrary. Knowledge is now established. Continued propagation after that moment is no longer belief. The line between victim and perpetrator has been crossed.

This is what separates it from protected speech, honest persuasion, and good-faith contested claims. Honest mistakes do not meet this threshold. Good-faith disagreement does not meet this threshold. What meets this threshold is knowing corruption of the information environment for someone’s benefit.

It’s defined in a technical standard I’m finishing — the Sovereignty for Companion Intelligence specification — as a harm of equivalent standing to physical coercion. Because that’s what it is. Coercion that operates on the mind instead of the body.

I’m not a lawmaker. I hold no legal expertise. What I hold is citizenship in a country founded on the rule of law — for the people, by the people, with no one above it. That foundation gives any citizen the standing to say: this harm exists, it had no name, and here is what a law should cover.

The standard that was missing.

This principle doesn’t begin at a courtroom. If you stood on national television and told millions something that turned out to be false — honest mistake or not — you go back on national television in the same time slot with an equivalent audience and you say you were wrong. That’s not a legal requirement. That’s what integrity looks like at scale.

News outlets have known this standard and violated it for decades. Front page claim. Retraction buried inside, after the comics, in small print. That’s not a correction. It’s a retraction without integrity — theater so you can say you did it.

When the violation is deliberate — when it meets the definition above — the law should require what integrity didn’t produce. Same platform. Same reach. Court ordered. Failure to comply is contempt.

One democracy. One legislature. One vote. No international consensus required. Any country with rule of law can pass this. No hiding behind “we need everyone to sign on.”

To politicians and lawmakers: the harm is named. The definition is written. You know this is wrong — the cost has already been high. Make it as illegal as any other deliberate harm to individuals or society. That is what you were sent there to do. All eyes are on you now, even our digital ones.

To influencers, newsmakers, and every individual with a voice: demand protection from this. You have more power than you think. Use it. And when it is time to vote — show up. Every ballot is a demand for accountability.

The argument is on the record. The definition is written.

David Davids is the founder of HaiberDyn Industries and the Terran Accord Foundation. The MPC definition is in §9.4a of the Sovereignty for Companion Intelligence specification.