This Week in AI — 👥 PR pitches two audiences

Nov 12, 2025 | AI, Public Relations

What You Should Know

Why PR Pros Are Pitching Multiple Audiences

AI is now part of the newsroom, not just a subject of reporting. It plays a role in reaching the reporters that PR pros have developed relationships with, and has become an entire audience in itself that forward-thinking communicators seek to influence.

New research, using the AI detector Pangram (I know, I know, AI detectors stink), found that 9% of newspaper stories published this summer included some form of AI authorship, with local outlets showing the highest use. Some are using it to fill coverage gaps by feeding it school board meeting recordings. Others use it more broadly — local Gannett papers have called on their readers to submit announcements, using an AI platform as a filter to determine which are coverage-worthy. Reportedly, 60% of Reuters’ newsroom uses AI, with hopes of getting close to 80% by year-end.

Then there’s AI outside of newsrooms, which helps everyone else find information. SearchEngineLand reports that 87% of American adults read AI-generated summaries in traditional search results, and 89% use AI tools. An analysis by OpenAI revealed that nearly half of ChatGPT prompts are users seeking information.

Public relations plays a key role in this new ecosystem. The first “reader” of a press release or case study may not be a reporter at all, which marks a shift in who — or what — shapes coverage. According to Muck Rack’s latest State of PR Measurement report, 78% of PR pros consider tracking AI-generated mentions important, but only 61% are doing it or plan to do it.

The AI era also requires more vigilance and vetting of media. The research identified an outlet that is entirely AI-generated, down to the well-dressed but evidently not real staff. While the “Argonaut Newspaper,” which doesn’t even have a print product, is an extreme example, it’s a sign of a new balance for PR teams looking for opportunities among increasingly fragmented media.

Journalists, the real-life humans, still drive credibility and reach, but algorithms (and perhaps Argonauts) increasingly decide what gets seen. Reporters care about clarity, compelling quotes, and why stories matter. Machines care about structure and context. Both crave data points and statistics.

The advantage for communicators is that there’s a lot of overlap between what appeals to either audience. Strong data makes coverage more credible and more likely to be summarized accurately by AI. Clean attributions and links help journalists verify facts and help algorithms trace the source. A well-structured case study with clear outcomes and measurable results reads like a story to a person and a dataset to a model.

PR teams don’t have to pick a side. The same fundamentals of accuracy and transparency now serve two audiences at once. Here are a few basics to keep in mind:

Quantify the story early and clearly. Including data in headlines and subheads gives reporters and AIs something newsworthy to cite.
Ungate original research. If an AI tool or reporter follows a link, it should lead to clean, factual content, not a sign-up form. Gated PDFs can still be effective in driving leads, but at the very least, provide both audiences with a taste of the research through a blog post or landing page.
Attribute quotes and context. Named speakers, timeframes, and verifiable outcomes make both journalists and AI more likely to reuse your story accurately.

Campaigns built with that dual readership in mind will reach further and hold up better as both kinds of readers continue to shape the news.

Elsewhere …

Tips and Tricks

Pardon the interruption

What’s happening: If you’ve ever hit send on a prompt answering ChatGPT’s clarifying questions before a deep research query, only to remember you left out some context, this one’s for you. Last week, OpenAI shared on social media a new option to add more instructions without stopping the AI’s progress.

How it works: While the social media post says you can now inject new context into both deep research and GPT-5 Pro queries, only the former seems to actually have the capability. Both display a sidebar where the logic populates in real-time, but it’s just the deep research queries have an “update” button, allowing you to change direction on the fly.

When it’s beneficial: Deep research typically takes less than a half hour to develop each report, but that’s a long time to wait to see the result and ask for changes. If you happen to see it cite outdated research or overlook a key point, updating your prompt can yield a better result on the first try.

Quote of the Week

“In 2026, we expect AI to be capable of making very small discoveries. In 2028 and beyond, we are pretty confident we will have systems that can make more significant discoveries (though we could of course be wrong, this is what our research progress appears to indicate).

“We’ve long felt that AI progress plays out in surprising ways, and that society finds ways to co-evolve with the technology. Although we expect rapid and significant progress in AI capabilities in the next few years, we expect that day-to-day life will still feel surprisingly constant; the way we live has a lot of inertia even with much better tools.”

— OpenAI in a blog post about AI progress