Picture a prospective customer, partner or donor sitting down to learn about your organization. A few years ago, they would have typed your name into a search engine and clicked on whatever came up first. Today, a growing number of them will ask an AI tool and read the answer it gives.
Where does that answer come from? A common assumption is that it comes from your website. Usually, it doesn't. AI tools lean heavily on what other people have said about you, and that has real consequences for how leaders think about visibility, reputation and strategy.
Most of what AI cites is earned media
About 84% of AI citations come from earned media, according to Muck Rack's research, as reported by PRSA's PRsay. In the same research, journalism accounts for 27% of all citations, compared with 13.7% from owned sources like company websites and social channels.
Earned media here is broader than news coverage. It also covers academic research, government agencies, nonprofits, encyclopedia sites and other third-party content that mentions you.
Your website still matters, because it's where you control the facts. But the evidence suggests AI tools build their picture of your organization mostly from what others report about it.
It isn't only the national outlets
Many leaders assume only big, well-known publications shape these answers. The data points to a more mixed picture. Featured, the company behind the HARO sourcing platform, analyzed more than 22,000 AI citations and found that about 34% came from websites with a domain authority below 40, and 76% of cited domains appeared only once. The full analysis is summarized by Spin Sucks.
Domain authority is a rough measure of how established a site is. In plain terms, a large share of what AI cites comes from smaller, more specialized sources that rarely make a traditional media list.
That doesn't mean national coverage has lost its value. Sites with very high authority still account for a substantial share of citations, so the realistic takeaway is both, not one or the other.
It also raises a question about where PR effort goes. Muck Rack's Natan Edelsburg said only about 2% of the journalists PR teams typically pitch are cited by AI, partly because many major outlets sit behind paywalls these tools can't easily read.
The research doesn't single out any one region or industry. But if your customers rely on a trade publication, a regional outlet or an association newsletter, coverage there may do more work than you've been giving it credit for.
Why this is a strategy question, not a PR tactic
Coverage can only reflect what an organization is clear about. Reporters, partners and customers describe you in their own words, but they draw on what you consistently say and do. If your leadership team would describe the company several different ways, the coverage will likely follow, and so will the answer an AI tool gives.
That's why the work starts before any pitch goes out. What do we want to be known for? Who needs to hear it? Which sources do those people already trust? These are leadership questions, and the answers make every placement more useful.
It's also why one-off launches do less for visibility than leaders hope. A single announcement is one moment. A steady record of specific, useful coverage built over months is more likely to give both people and AI tools a consistent story to repeat.
What leaders can do now
Most of it starts with paying attention.
Ask the tools directly. Pose the questions a customer, partner or donor would ask about your company and your category in more than one AI tool, since they don't all cite the same sources. Note what they get right, what's missing and which sources they point to. Check again every few weeks, because the answers shift, and ask new customers how they found you.
Map the sources that matter. From those answers, list the publications that keep appearing in your industry and compare them with where you currently earn coverage. Include trade publications, regional outlets and association newsletters alongside the national names.
Build a steady rhythm instead of a spike. Plan specific, useful stories over months, such as customer stories and commentary from your own experts, rather than one announcement. Write your own pitches, too: in Featured's HARO data, human-written pitches were three times more successful than AI-written ones.
Keep your own facts consistent. Your website, leadership bios and organization descriptions are still where you control the record. Make sure they say what you want others to repeat.
The same trust, with a new audience
Reputation has always been built by what others say when you're not in the room. AI hasn't changed what earns trust. It has added a new audience that reads the same record, and it has made it easier to see who has been paying attention.
For leaders, the practical message is simple. Decide what you want to be known for, say it consistently, and make it easy for credible third parties to repeat. The coverage that results serves the people reading it today and the tools answering questions about you tomorrow.
Sources
AI Citations Spark 'Renaissance Moment for PR and Communications', PRsay (PRSA), October 1, 2026. Source of the Muck Rack figures on earned media, journalism and owned media, and the comments on pitching.
Are PR Teams Doing Earned Media Wrong in the AI Era?, Spin Sucks, October 15, 2026. Source of the Featured findings on domain authority, the long tail of cited sources, and human-written pitches.
