There’s a new kind of cold email hitting journalists’ inboxes, and it’s somehow even less effective than the old kind.
We already talked about the cold communicator — the PR approach that treats outreach like a numbers game and reporters like names on a list. AI-generated pitching is that problem, but industrialized. The volume has gotten so absurd that reporters at publications across the tech beat have started publicly posting examples: pitches that reference the wrong publication, compliment journalists on articles they didn’t write, or open with phrases so generic they could have been addressed to literally anyone.
Because they were.
The Tell Is In the Details (Or the Lack of Them)
Most AI-generated pitches share a few unmistakable traits. They’re polished in a way that feels oddly formal, full sentences, no typos, perfectly structured, but completely devoid of anything specific. There’s no mention of a recent article the reporter actually wrote. There’s no acknowledgment of their particular beat or angle. There’s nothing that says I know what you cover, and this story fits it.
Instead, you get something like: “As someone who covers the intersection of technology and business, I thought your readers might be interested in…” A line that could be copy-pasted into a pitch to 500 reporters without changing a single word.
Journalists are not fooled by polish. They’re looking for relevance. And a pitch that reads like it was assembled from a template signals immediately that the sender didn’t actually think about them.
What You’re Communicating Without Realizing It
The problem with AI pitches isn’t just that they’re generic. It’s what they say about how you see the reporter on the receiving end.
When a journalist receives a pitch that’s clearly been bulk-generated, the message it sends is: I couldn’t be bothered to learn anything about you or your work, but I’d like something from you. That’s not a great starting point for a relationship you’re going to need down the line.
Reporters talk to each other. They share bad pitches. They remember the companies that wasted their time, and they remember the PR contacts who seem to actually understand journalism versus the ones who are trying to hit a quota. That reputation, good or bad, follows your company around.
The Efficiency Math Doesn’t Work
The appeal of AI pitching is obvious. You can generate 200 personalized-sounding outreach emails in an hour. Why spend three hours researching and crafting five thoughtful ones?
Here’s why: a 0.1% response rate on 200 lazy pitches is worse than a 40% response rate on five well-researched ones, both in coverage outcomes and in the long-term cost to your credibility. The journalist who ignores your AI pitch today is the same one you’ll need to pitch again in six months when you actually have something newsworthy. That first impression is already made.
And in a media environment where newsrooms are leaner than ever and reporters are stretched across more beats, their patience for low-effort outreach is at an all-time low. They’ve gotten very good at deleting fast.
What Good Pitching Still Requires
None of this means AI tools have no place in PR. They’re genuinely useful for drafting, editing, and thinking through angles, behind-the-scenes work that can help a human get to a better pitch faster. That’s a reasonable use of the technology.
But the pitch itself, the thing that lands in a journalist’s inbox and asks for their time and attention, still needs to reflect actual thought. That means reading their recent work. It means having a concrete reason to contact this specific reporter about this specific story. It means writing something that would only make sense to send to them, not to a list.
That kind of specificity isn’t something you can automate away without a human checking the work carefully enough to largely defeat the purpose. The reporters worth pitching are smart enough to know the difference. Act accordingly.