The Movchan Agency Scandal Doesn't Mean What You Think It Means About AI in PR

Last week, Futurism broke a story that sent the PR world into a frenzy: a UK-based "viral PR" firm called Movchan Agency had been caught operating at least 15 fake PR personas, complete with AI-generated headshots, fabricated names, and dummy company domains, to pitch journalists on behalf of paying clients.

The journalist who broke the story, Maggie Harrison Dupré, noticed a watermark from this-person-does-not-exist.com on a sender's headshot. One investigation later, the entire scheme unraveled. The fake personas traced back to Movchan. The clients had no idea. The industry erupted in outrage.

And almost everyone is drawing the wrong conclusion.

The wrong lesson

The dominant takeaway from the Movchan story has been: AI agents are dangerous in PR. Fake personas, automated spam, deception. This is what happens when you let machines do a human job.

That's a comforting narrative. It's also wrong.

Movchan didn't fail because they used AI. They failed because they lied and because they treated PR as a volume game. Neither of those problems is caused by technology. Both are caused by strategy, or the absence of one.

Problem one: the lie

Let's be clear about what Movchan actually did. They didn't just use AI tools to help with their work. They fabricated human beings. Fifteen of them. Each had a fake name, a fake email address at a fake company domain, and in several cases, an AI-generated face downloaded from a website whose entire purpose is generating faces of people who don't exist.

They hid this from journalists. They hid it from their own clients. When a founder at one of their client companies was contacted, he said he had "no idea" the firm was using fake personas and that it "goes entirely against our brand values."

This isn't an AI problem. This is an honesty problem.

If you're using AI agents in your PR outreach, whether for drafting, sending, or managing pitches, there is a simple, obvious solution to the ethical concern: disclose it. Include a line in your email signature. Add a note to your pitch. Say "this message was sent with the assistance of an AI tool" or "our team uses AI agents to manage outreach." Journalists are not stupid. Most of them won't care. What they care about is whether the pitch is relevant and whether the person on the other end is reachable when they want to follow up.

And if you get caught doing something you didn't disclose, admit it. Movchan's founder claimed the fake personas were a "temporary" fix for a domain reputation problem, purchased as "pre-warmed" accounts from a vendor. But journalists found fake pitches dating back almost a year. The excuse didn't hold, and the cover-up made it worse.

The lesson here isn't "don't use AI." It's "don't lie."

Problem two: the numbers game

The bigger issue, and the one almost nobody is talking about, is that Movchan was running a spam operation dressed up as PR.

Think about what happened: a single PR firm was operating 15 fake personas across 15 fake company domains, all pitching the same journalists about the same clients. That's not targeted media relations. That's carpet bombing. The fake personas weren't created to improve pitch quality. They were created to increase pitch volume.

This approach was broken long before AI agents existed. PR firms have been buying journalist email lists and blasting generic pitches for decades. The hit rate is low, the relationships are nonexistent, and the only thing it produces is a number you can put on a monthly report: "We sent 500 pitches this month."

That number means nothing. What matters is whether a journalist actually wrote a story because of your pitch. And the thing that makes a journalist write a story is a combination of: (1) the story is genuinely newsworthy, (2) the pitch is relevant to their beat, and (3) they trust the person pitching them.

None of that improves with volume. All of it improves with relationships.

You can't automate your way out of a strategy that doesn't work. If your PR model is "send more pitches, hope for the best," AI agents will just help you fail faster.

What AI agents actually do well in PR

Now for the part the outrage crowd is missing. AI agents have genuinely valuable applications in PR, applications that, done right, benefit both PR firms and journalists.

Automating the mechanical work

Sending a pitch is a manual process. You write the pitch in a document. You open your email client. You copy the subject line. You paste the body. You type in the journalist's address. You hit send. Then you do it again, 30 times, with slightly different personalization for each recipient.

That process is tedious. It's also completely mechanical. Automating it, having an AI agent pull a pitch from a Google Doc, populate the subject and recipient fields, and send the email, doesn't reduce pitch quality at all. The pitch is still written by a human. The personalization is still done by a human. The strategy is still set by a human. The only thing that changes is that the human isn't spending two hours copying and pasting.

That's not deception. That's efficiency. And any journalist who's upset about a well-written, relevant pitch arriving via automation instead of manual copy-paste is performing, not reasoning.

Assisting with pitch writing

AI can help draft pitches today, but let's be honest about where the technology actually stands. If you've used AI to write a pitch, you know it's not a one-shot process. You generate a draft, read it, spot the problems, ask for a rewrite, spot new problems, ask for another rewrite. You might go through 6 or 7 drafts before you land on something a journalist would actually want to read. That's still faster than starting from a blank page, and a good PR professional can shape the output faster than they could write it from scratch. But it's not hands-off.

The future we're heading toward is one where most pitches come back good enough to approve without edits, only needing a human's touch every once in a while. We're not there yet. But the gap between "AI as a drafting tool that needs heavy editing" and "AI as a pitching tool that just needs a quick review" is closing faster than most people imagine.

The key word in both the present and the future version is "oversight." A human reviews every pitch before it goes out. A human edits for tone, relevance, and accuracy. The AI is a tool, not an author. Used this way, it doesn't lower quality. It raises the floor while keeping the ceiling where the human puts it.

Newsjacking: the use case nobody's talking about

This is where AI agents get genuinely exciting.

Newsjacking, inserting your client's commentary into a breaking news story, is one of the highest-value tactics in PR. But it's also one of the hardest to execute well, because the window is short and the research is time-consuming.

An AI agent can change that. Here's how the workflow could work:

1. The agent monitors news feeds in real time across dozens of outlets.
2. It flags stories that one publication has covered but other publications haven't picked up yet.
3. It identifies journalists at those other publications whose beats include the topic.
4. It recommends which journalists to pitch and provides context on what angle they'd likely find interesting.
5. The PR team reviews the recommendation, crafts a pitch offering their client's exclusive commentary, and sends it.

This doesn't just benefit the PR firm. It benefits the journalist. They learn about a story they might have missed. They get exclusive commentary that makes their coverage stand out from competitors. They get a source they can quote, a real source, with real expertise, offered transparently.

Compare that to what Movchan did: blasting generic pitches from fake identities at journalists who never opted in. The technology is the same. The strategy is the opposite.

Where this is going

We may be approaching a future where AI agents handle the full pitching process: monitoring news, identifying opportunities, writing pitches, and sending them, all under human supervision. The human's role shifts from "doing the mechanical work" to "making sure the mechanical work is done well."

But one role will always stay human: the client relationship. An AI agent can pitch a journalist. It can't sit on a strategy call, understand a founder's anxiety about a token launch, or reassure a client that the coverage is coming. It can't read the room in a pitch meeting and decide to pivot the angle. The account manager role, the relationship role, is human. And it's going to stay human.

The real lesson

The Movchan scandal is a story about a PR firm that lied to journalists and clients, spammed inboxes with fake identities, and got caught. The outrage is justified. The firm deserves every bit of it.

But the lesson isn't "AI is bad for PR." The lesson is: be honest about what you're using, don't treat PR as a volume game, and build relationships instead of buying email lists.

The firms that learn to use AI agents transparently and strategically, to automate the mechanical, to assist with the creative, and to surface opportunities humans would miss, are going to have an enormous advantage. Not because they're cutting corners, but because they're spending their human time on the things that actually require a human.

The firms that shout "AI is wrong" and keep doing everything by hand are going to be outpaced by the ones that figure this out.

The future of PR isn't fake personas and spam. It's real relationships, smarter tools, and the honesty to tell people which is which.

Movchan got that backwards. You don't have to.