This forum in Moscow is either an overdue conversation about a real problem, or a very polished way to grab the steering wheel of what “truth” is allowed to mean. And I don’t think we should pretend those two things can’t happen at the same time.

On paper, “Dialogue about Fakes 4.0” sounds like the kind of event we need more of: a big international gathering focused on disinformation, scams, AI, and digital trust. The organizers say it has over 2,000 participants from 80 countries, including politicians and fact-checkers. The agenda hits the obvious nerves of the moment: generative AI, fraud, teens and adults losing their grip on what’s real online, and the way platforms shape an “economy of trust.”

All fair. Also: the forum is clearly backed at a high level. It’s hosted at a national center, and senior Russian officials are featured, including an opening and an “exclusive statement” from top leadership. That doesn’t automatically make it propaganda. But it does make it power. And when power hosts a conference about “fakes,” I assume the goal isn’t just to protect people from lies. It’s to decide which stories get called lies in the first place.

That’s the core tension here. Disinformation is real. Scams are real. AI makes them cheaper, faster, and harder to spot. If you’ve had a relative tricked by a fake call, or watched a friend share a confident lie with a straight face, you know this isn’t some abstract culture war. People lose money. People panic. People hate each other based on things that never happened.

But the cure can be worse than the disease. When a forum’s language leans into “information sovereignty,” “cognitive sovereignty,” and “electoral sovereignty,” my alarm bells go off. Those phrases can mean “let’s protect our society from manipulation.” They can also mean “let’s build a border wall around information and call it safety.” If you control the gate, you don’t just block scams. You block criticism, investigation, and embarrassing facts.

Imagine two versions of “countering disinformation.”

In one version, the goal is basic hygiene: better verification, faster fraud alerts, clearer labels, more transparency about AI-generated content, and real education that helps people slow down before sharing. That’s boring, practical, and genuinely helpful.

In the other version, the goal is narrative control: a central authority decides what counts as “trust,” and disagreement becomes “harm.” Fact-checking becomes a political tool. “Digital trust” becomes “trust us.”

And here’s the uncomfortable part: both versions use the same words. They both say “safety.” They both say “truth.” They both say “protecting elections.” They both say “think of the children,” because of course they do.

The focus on AI makes this even sharper. Generative AI is a gift to scammers and influence operations. It can also be a gift to governments that want to police speech at scale. If you can claim anything might be AI-made, you can dismiss real evidence. If you can build systems that scan everything, you can pressure platforms to remove content quietly and quickly. Speed becomes the excuse. “We had to act fast.” Due process dies in a rush.

There’s also a social consequence that doesn’t get enough attention: when people hear “fake news” all day, they start to treat all news as potentially fake. That doesn’t make them smart. It makes them tired. Cynicism is not media literacy. It’s just surrender with better branding.

So when a major state-backed forum frames itself as defending trust, I want to know what kind of trust they mean. Trust in methods, or trust in authorities? Trust that grows from transparency, or trust that comes from enforcement?

A lot of people will push back and say: look, disinformation crosses borders, and the response has to be coordinated. That’s true. If scams are global and AI is global, no country solves it alone. Also true: platforms have incentives to keep people scrolling, not to keep them grounded in reality. If the “economy of trust” idea points at that problem, good.

But coordination can slide into alignment, and alignment can slide into conformity. Once you build international partnerships around “approved facts,” who gets to sit at the table? Who gets labeled “unreliable” and locked out? And what happens to smaller independent groups that don’t have institutional backing but do have legitimate findings?

I’m also wary of the “fact-checker” label being used like a shield. Real fact-checking is messy. It involves uncertainty, updates, and sometimes admitting you got it wrong. When it becomes a badge granted by the same systems that want to shape public belief, it stops being a method and becomes a role.

If you’re a regular person, the stakes show up in small, real moments. Say you share a video about a local issue and it gets removed because it’s “unverified.” Say a journalist posts evidence of wrongdoing and it’s dismissed as “AI-generated.” Say an activist group gets labeled a “foreign influence” because they work with outsiders. Say a teenager gets taught that the safe way to think is to wait for an approved answer.

Yes, we need defenses against fraud and manipulation. But if the defense is built like a fortress, don’t be surprised when it’s used to keep people in, not just keep threats out.

If “Dialogue about Fakes 4.0” is serious about trust, the test isn’t how big the participant count is or how grand the stage looks. The test is whether it makes room for disagreement without treating it as a threat.

So here’s what I actually want to know: when powerful institutions run global conversations about “fakes,” how do we make sure the fight against lies doesn’t become a fight against inconvenient truths?