Recently I sent the email every PR person loves sending, “Published!” But this time, the client followed up with a question that just recently came into existence: “Did this affect our AI search?” 

That second question is part of the job now. PR’s core discipline has always been earning the attention and trust of journalists, and that hasn’t changed. Over the last couple of years, a second audience showed up. It’s one that doesn’t have a beat and doesn’t write.

Every AI-powered engine now pulls from the same pool of published coverage we’ve fought to earn, and it uses that coverage to describe our clients to anyone who asks. PR leaders are no longer just managing our relationships with reporters. We’re managing what the algorithms remember.

The higher stakes

Relationships in media relations are still so important and that’s not being automated away. Check on LinkedIn and plenty of PR professionals and journalists themselves say so. Storytelling and content creation remain the single most in-demand skill in the profession in 2026, with media relations close behind. The instinct, judgment and trust-building that go into a good pitch are still entirely human work.

What’s different isn’t whether a story has a long life — good coverage has always had one, whether through SEO, repurposing into sales collateral or a reporter’s own archive resurfacing in later stories. Now it’s about who’s reading that story with a long life. In search-engines it meant a human eventually found the story and drew their own conclusions. But now AI models are synthesizing stories into flat, declarative answers, handing them to anyone who asks, without the reader seeing the original piece or receiving potentially important context. 

That’s a real shift in stakes, even if our day-to-day looks the same.

Monitoring for two audiences

Monitoring in PR has typically meant tracking metrics like mentions, readership and share of voice. It was making sure coverage hit, was accurate and got in front of the right audiences. That work is still essential. But there are plenty of nuances that make it harder than it may look from the outside: following up with reporters who went dark after an interview or editors ghosting after you send them a byline, plus staying on top of a beat among what feels like constant news organization layoffs requires real discipline, not just a dashboard.

Add a second audience that never sleeps and never forgets, and now that discipline matters even more. Answer Engine Optimization, or making sure a brand shows up accurately and favorably when someone asks an AI system a question instead of Googling it, is no longer a buzzword. It has become a real priority for communicators this year. Adoption numbers back up how fast this has moved from experimental to expected. A significant share of PR professionals now use AI-driven media monitoring tools, and a growing number rely on AI-powered reporting to track how coverage performs, which isn’t just saying whether it ran, but also how it’s being interpreted downstream.

The demand on PR teams

I’m not saying to throw out what works, but we do need to be more rigorous about it. There are a few shifts that can be made:

  • Redefine what “good coverage” means: The quantity of hits matters less than it used to. Because AI systems weigh credibility and consistency, a smaller number of accurate, well-placed stories can outperform a dozen thin ones.
  • Build monitoring that serves both audiences from the start. Tracking coverage for a client update and tracking how AI represents a brand shouldn’t be two disconnected efforts. They’re the same discipline, applied to two outputs.
  • Treat consistency as the differentiator. Sloppy or outdated coverage used to be a minor liability. Now it could live in perpetuity. The teams that win are the ones who catch inaccuracies early and keep the record straight, not just the ones with the flashiest campaign ideas.

Keeping the fundamentals the same

Good PR has always come down to earning trust with reporters, with a target audience and with a client. That’s not going to change. What has changed is that the coverage we earn now has to hold up for two audiences at once: the human one that reads it today, and the algorithmic one that will keep referencing it long after publication.

The teams that already run their monitoring and coverage processes with discipline and creativity — not just one or the other — are the ones best positioned to adapt. It’s not a talent gap, but an operational one, and it’s one worth closing now.