Each week we'll gather headlines and tips to keep you current with how generative AI affects PR and the world at large. If you have ideas on how to improve the newsletter, let us know!
What You Should Know
What Journalists’ AI Use Means for PR Teams
Journalists’ mixed feelings toward AI are well-documented in LinkedIn posts and media analysis articles. The new Cision 2026 State of the Media Report offers a more detailed picture. Cision found that 48% of journalists use generative AI for brainstorming, including story angles, interview questions, and headlines, while 43% use it for research or fact-checking, and 41% use it for transcription or summarization of interviews and audio.
Perceptions change when the AI output comes from PR. Cision found that 53% of journalists oppose PR teams using AI to generate pitches or press releases because of low-quality writing and a lack of personalization. The issue is not the tool itself, but what the tool often produces when no one brings judgment to the output.
AI can make a pitch sound clean, but clean is not the same as useful. It can summarize a company announcement, polish a quote, and turn a few bullets into a press release. None of that matters much if the result is generic, thinly sourced, irrelevant to the reporter, or missing a real story. Used well, AI can help PR teams get there. It can analyze an angle, flag vague language, summarize background, suggest sharper subject lines, and identify where a pitch needs more context. It can also help uncover the data and research that would get a journalist’s attention in the first place.
At the same time, 66% of reporters still rely on PR-provided content, including press releases, pitches, and media kits, for story ideas. The report also found that 79% of journalists are more likely to engage with a pitch that is relevant to their beat, audience, or coverage area.
The basics of good media relations have not changed. Relevance, accuracy, sourcing, and access have always mattered. The new wrinkle is that journalists now have AI in their own workflows, which means they have less patience for PR materials that use AI as a shortcut around those basics.
Before sending AI-assisted materials, PR teams should be able to answer a few questions. Is the pitch actually relevant to this reporter? Are the claims sourced? Is the data current? Is there a clear reason this matters now? Does the quote sound like something a person would actually say? Is the expert available and prepared? Can the journalist quickly verify the most important points?
That last question carries more weight now. When journalists are already balancing AI adoption, misinformation concerns, and resource pressure, verification is part of the service PR can provide.
Elsewhere …
- LISTEN: Advisors Are Saving 10 Hours a Week with AI — and Wasting Every One of Them
- Tech Giants Back New Data Center Climate Initiative
- Robinhood Opens Platform to AI Agents for Stock Trading and Credit Card Spending
- AI Search Is Pushing 44.8% of Brands to Spend More on PR
- New USGA Tool Uses AI to Settle Your Golf Rules Debates
Tips and Tricks
🗂️ Curate your context
What’s happening: AI agents need context to do useful work, but more information does not always mean better output. If you give an agent every call transcript, old strategy doc, performance report, pitch draft, and meeting note you can find, you may actually make the work worse. The model has to sort through everything, and what it really needs is to know what “good” looks like.
Why it matters: Overloading context leads to messy answers. If an agent pulls from outdated messaging, overweighs last month’s campaign, or mimics posts that no longer match the brand, the source material has already soured the output.
How to do it better: Treat context like an assignment editor would. Decide what the agent should always know, what it should ignore, and where a person needs to review the work before it moves forward. Make sure reference files for messaging, audience priorities, and client-approved language are all current and strong examples. Giving the agent the right material, clearly labeled, with enough human judgment around it, will make outputs more useful.
Quote of the Week
“I am accomplishing so much more that I couldn’t have accomplished before. It’s probably a 50% time savings on the weeks it’s deployed.”
— Reid Hoffman, Co-Founder of LinkedIn and Partner at venture firm Greylock Partners, to the Wall Street Journal on his digital twin
How Was This Newsletter?
