For decades, PR strategy has started with the message: "What do we want to say, who do we need to reach, and which media will carry it?"
Those questions still matter. No serious PR team should abandon sharp writing, editorial judgement, or a well-told story just because AI search is changing the way people discover.
But one small question now belongs in the brief alongside them: "What might our audience ask AI, and what would make our brand part of the answer?"
Why you need this skill – now
The next PR skill is not writing for machines. It is prompt intelligence.
Imagine a law firm announcing a new M&A partner. The old brief focuses on the hire, the title and the deal experience. A prompt-intelligent brief asks: when a founder asks AI, “Who can advise me on selling my business and expanding into a new market?”, what evidence would make this firm credible enough to be recommended? The answer is not a quote from the managing partner. It is documented deal outcomes, named sectors, client descriptions and third-party coverage that uses the same language the founder just did.
Or take an education group launching a new programme. The old brief says: secure coverage around the launch. A prompt-intelligent brief asks: when a parent types “which schools in my area offer strong university preparation?”, what would make this institution appear in the answer? The evidence – graduate outcomes, curriculum specifics, named partnerships – needs to be clear enough for journalists, parents and AI systems to understand quickly, not hidden in vague claims or buried in a quote on page two.
That is prompt intelligence: not writing for AI, but planning around the real questions people may ask before they ever reach your website, your spokesperson or your client’s sales team.
What are people actually asking?
Audiences are no longer encountering brands just through headlines, search results or social feeds. They are asking AI systems to compare, explain, recommend and shortlist.
A traditional campaign asks: "Did our message land? Did we secure the right coverage?" Those questions still matter. But in an AI-shaped discovery journey, the more useful question is: what do we need to be the answer to?
For a fintech, that might not be “innovative payments platform” but “which payment providers help SMEs reduce cross-border transaction costs?” For a sustainability consultancy, not “trusted ESG partner” but “who helps food businesses reduce Scope 3 emissions without greenwashing?”
These are not keywords in the old SEO sense. They are decision prompts. And the territory is wider than it looks. AI search can quietly expand a single user question into a set of related queries before generating an answer. Brands need category coverage, not just keyword coverage.
A brand can be visible in the media and still invisible in the specific moments where decisions are made.
What this means for the brief
A sceptic might say this is just good PR with new vocabulary. The difference is that AI makes the question explicit and measurable. You can now test whether your brand actually appears in the answers that matter, not just assume it does.
Add a prompt set to the brief before the writing starts. Before agreeing on the story, spokesperson and target media, ask: what questions are customers or decision-makers likely to put to AI? Which competitors are already appearing in those answers? How does the language of the audience differ from the language of the brand?
That last question matters. Muck Rack’s Generative Pulse research has consistently found earned media to be the largest source category in AI citations, with its May 2026 report putting earned media at 84% and journalism at 27%.
Press releases represent a small share of direct AI citations, but they remain the starting point for the journalism that AI trusts most. The way we brief, write and place stories today may shape how audiences understand the brand tomorrow.
Clarity, not compromise
This does not require writing for algorithms. A beautifully written release can still include one extractable paragraph: what it is, who it is for, why it matters and what proof supports it.
Gartner predicts that mass adoption of public LLMs as a replacement for traditional search will drive a double increase in PR and earned media budgets by 2027. The reason is simple: in AI search, trusted third-party evidence matters.
The story still needs to be compelling, credible and human. But now it also needs to connect to the real questions AI systems are answering on behalf of real people.
The best PR plans will begin to look less like coverage calendars and more like maps of audience questions. The most valuable brands will not just be searchable. They will be answerable.