If your pitch isn’t getting through in 2026, it’s not because journalists are ignoring you. It’s because AI filtered you out before they even saw it.
This week, Cleveland’s News5 published their AI policy: “Enhance, not create.” Ideastream Public Media stated they “will only consider using AI that expands human capabilities over automation that could displace jobs.” The AP automated their earnings reports and saw a 400% output increase with fewer factual errors.
The pattern is clear: AI is now embedded in newsroom infrastructure. It’s not a future trend — it’s operational reality. And if you don’t understand how it works, your pitches are dying in the filter before they reach a human.
Here’s the good news: AI newsroom tools are predictable. They’re looking for specific signals, and once you know what they are, you can structure your pitches to pass the filter and get to the journalist.
This isn’t theory. I’ve spent the last six months analyzing newsroom AI policies, testing pitch formats, and tracking what gets through. Here’s the tactical playbook.
Let me clear up a misconception: AI isn’t replacing journalists. It’s filtering the noise so journalists can focus on stories that matter.
Think about it from a reporter’s perspective. They get 200+ pitches a day. 90% are irrelevant, poorly timed, or clearly mass-blasted. A human can’t read all 200. So newsrooms are using AI to:
According to Editor & Publisher, “Journalism is at an AI inflection point” — and newsrooms are taking a hands-on approach, testing tools instead of following hype.
The result? Pitches that used to get opened 20% of the time now get filtered at 80% if they don’t meet AI’s criteria. But pitches that do meet the criteria? They’re landing with journalists who are pre-sold on the story’s relevance.
Here’s what I’ve learned from analyzing newsroom AI policies and testing pitch formats:
AI tools scan for time-sensitive hooks. If your pitch is about a trend that happened three months ago, it gets deprioritized. If it’s about something happening this week or tied to breaking news, it gets surfaced.
What to do: Lead with the time hook in your subject line and first sentence.
AI knows when reports drop, when earnings are released, and when industry events happen. Tie your pitch to something time-bound and you’ll pass the filter.
AI tools cross-reference claims against known data sources. If you say “most companies struggle with X” but don’t provide a source or number, AI flags it as vague. If you say “75% of marketers report measurement challenges, per IAB’s State of Data 2026 report,” AI verifies the source and prioritizes the pitch.
What to do: Include specific numbers and cite the source in the pitch body.
The more specific and verifiable your data, the higher your relevance score.
AI tools analyze a journalist’s past articles and prioritize pitches that match their beat. If you’re pitching a cybersecurity story to a reporter who covers healthcare, AI filters it out.
What to do: Research the journalist’s last 10 articles before pitching. Mention a recent piece they wrote and explain how your story builds on it.
AI sees the connection to their beat and surfaces the pitch.
Newsrooms are overwhelmed. AI tools prioritize pitches that are ready to become stories with minimal rewriting. That means:
What to do: Write your pitch like a press release that a journalist could turn into an article with 20% editing.
| ✅ Good: “Headline: ‘Newsrooms Are Using AI to Filter Pitches — Here’s How PR Teams Can Adapt’ | Key Quote: ‘AI isn’t replacing journalists, it’s filtering the noise so they can focus on what matters,’ says Christian Lehman, AuthorityTech. Data: 72% of journalists now use AI in their workflow (up from 18% in 2024). Available: chart showing AI adoption by newsroom size.” |
AI sees a structured story and flags it as high-priority.
Here’s the paradox: AI tools prioritize pitches that feel human. Generic, template-blasted emails get filtered out. Personalized pitches that reference the journalist’s work, ask a real question, or offer exclusive access get surfaced.
According to World Governments Summit 2026 coverage, media leaders emphasized: “We absolutely cannot at this moment in time replace the human connect that reporters bring to the newsroom by telling the story with empathy.”
AI tools know this. They’re trained to detect personalization signals: the journalist’s name used naturally, reference to their past work, a specific question that shows you read their coverage.
What to do: Spend 2 minutes researching the journalist before you hit send. Mention something specific.
AI sees the personalization and prioritizes it.
Based on everything I’ve learned, here’s the structure that works in 2026:
Subject Line: [Time Hook] + [Specific Data] + [Relevance to Their Beat]
Example: “New IAB data: 75% of marketers can’t measure earned media — follows your Feb 2 piece on attribution”
Body:
Hi [Journalist Name],
I saw your [date] piece on [specific topic] — [one sentence about what struck you or a question it raised].
I have new data that builds on this: [specific stat with source]. [One more supporting stat].
Story Angle: [Clear headline or angle in one sentence]
Why Now: [Time hook — report just dropped, earnings released, industry event, regulatory change]
What’s Available:
Would you be interested in covering this? I can send the full report and connect you with [expert] for quotes.
Best,
[Your Name]
This structure hits all five AI signals: timeliness, data specificity, beat relevance, story structure, and human connection.
Let’s be clear: journalists aren’t hiding behind AI. They’re using it as a co-pilot.
Yahoo News Canada reports that “AI is used in our newsroom every day” — for identifying sources, double-checking facts, and surfacing story ideas.
That means:
If your pitch passes the filter, you’re landing with a journalist who’s already pre-sold on its relevance. If it doesn’t, you’re invisible.
Here’s the weird part: people want journalists to say when they use AI, but trust drops when they do.
That’s the paradox newsrooms are navigating. They’re using AI to filter pitches, check facts, and surface stories — but they know disclosure can hurt trust.
What does this mean for you? Don’t mention AI in your pitches. The journalist knows they’re using AI tools. You don’t need to acknowledge it. Just structure your pitch so it passes the filter.
Focus on the five signals. Write like you’re pitching a human (because you are — AI is just the first gatekeeper).
If newsrooms are using AI to filter pitches, here’s how to adapt:
Spend 2 minutes reading the journalist’s last 5 articles. If your story doesn’t connect to their beat, don’t pitch. AI will filter it out anyway.
“We’re excited to announce…” is dead. Lead with the time hook and the most compelling stat. AI prioritizes data-rich pitches.
AI extracts who/what/when/where/why. Make it easy. Put the key facts in the first paragraph, not buried in the third.
AI tools prioritize exclusive stories because they know journalists do. If you can offer first access to data, an interview, or a report, say so upfront.
If a pitch lands, reverse-engineer why. Was it the time hook? The data? The personalization? Use that pattern for the next one.
Most newsrooms are using custom-built AI filters integrated with their CRM systems (like Cision, Muck Rack, or Meltwater). Some use GPT-based summarization tools to extract key facts from pitches. A few are testing LLM-powered “pitch scoring” systems that rank incoming emails by relevance. The specific tools vary, but the criteria are consistent: timeliness, data specificity, beat relevance, story structure, and human connection.
No. Write for journalists. AI is filtering based on what journalists value: relevant stories, timely hooks, verifiable data, and human connection. If you write a great pitch for a journalist, it’ll pass the AI filter. If you write a generic, data-free pitch, both AI and the journalist will ignore it.
You can’t, directly. But if you’re getting zero opens on pitches that used to get 20% open rates, it’s likely an AI filter issue. Test the five-signal structure (timeliness, data, relevance, structure, personalization) and track whether your open rates improve. If they do, you were getting filtered. If they don’t, it’s a relevance or targeting issue.
Want to see how your brand shows up in AI search?
Newsrooms aren’t the only place AI is filtering content. AI search engines like ChatGPT, Perplexity, and Gemini decide which brands get recommended based on earned media authority.
Get your free visibility audit and see where your brand appears (or doesn’t) in AI-generated answers.
— Christian