The most powerful AI visibility strategy isn't an algorithm

When I published an analysis of The Corporate Affairs Divide, one section generated more response than any other. It was a single paragraph about AI visibility. Specifically, that a downgraded communications function is not just a reputational risk. It is an AI visibility risk.

The response told me two things. My corporate affairs peers already feel this problem. And nobody had yet said clearly enough what was causing it, or what was actually at stake.

The data moved fast

When I wrote that piece, Muck Rack's research put earned media at approximately half of all sources cited by AI-driven GPTs. Only a few weeks have passed.

Their updated Generative Pulse research puts the same figure at 84%. Multiple studies now sit between 85 and 95%. The most rigorous to date is Stacker's GEO research across 87 stories, 30 brands, and 8 AI platforms. It found that earned media distribution produces a median 239% lift in AI search visibility.

Distributed stories were 5.3 times more likely to be the sole source of a brand's AI visibility than their own website. In less than one month, the evidence had moved decisively.

What earned media actually is and why it matters

Earned media is third-party coverage, mentions, and references that an organisation does not control directly. In AI search, it matters because answer engines look for sources they can retrieve, compare, and cite without treating them as self-assertion. That is the structural difference.

An organisation can have an excellent website, a sophisticated content strategy, and significant investment in owned channels, and still be largely invisible in the AI-driven answers that customers, investors, regulators, journalists, and prospective employees are now using as a primary research tool. 2X Marketing's 2026 AI Visibility Index found that 96% of B2B brands are effectively invisible in AI-generated answers.

This is not a content problem. It is a credibility problem. And credibility, in the AI environment, is determined by what others have chosen to say about you - not what you have said about yourself.

The journalist nobody is talking about

The past six months have seen significant writing on this topic from Forbes, Fast Company, AdNews, and across LinkedIn. My observation is that almost all of it is focused on reaching brand marketers. The advice is not wrong, but it consistently misses the harder question: what happens to the organisations that no longer have the capability to act on it or never built it at all?

There is a mechanism at the centre of AI visibility that every article mentions but none examine closely.

The AP signed licensing agreements with OpenAI, Google, and Microsoft. The Financial Times, News Corp, The Atlantic, and Vox Media followed. These publications now sit in a privileged tier inside the retrieval systems that power the AI products your stakeholders use every day. Media coverage in licensed, high-authority publications is a categorically different asset class. Coverage breadth - appearing consistently across multiple outlets and platforms - is now the primary authority signal that AI systems use to determine who is worth citing.

What produces that media coverage is not only content strategy or GEO optimisation (which remain important). It is a journalist who trusts you enough to call you first. An editor who knows your organisation will provide substance rather than talking points. A reporter who has found, repeatedly, that your experts deliver on what they promise.

Those relationships are the product of years of investment in access and substance. This is the kind of credible engagement that makes a journalist choose you over the dozens of other organisations competing for the same space.

That relationship runs both ways. As newsrooms have contracted - fewer journalists, more generalists covering more ground - the communications professional who provides reliable access to genuine expertise, accurate data, and a story worth telling has become more valuable to journalism, not less. A trusted contact who understands what makes something newsworthy, who never wastes a journalist's time, and who delivers on what they promise is a resource that stretched reporters depend on. When organisations remove that function, they do not just lose access to journalists. They lose the ability to be useful to them and risk disappearing from the journalism ecosystem entirely.

The communications function that built and maintained those relationships was not just doing reputation work. It was building one of the most important AI visibility assets in the organisation.

The double penalty

In my previous piece, I wrote about the split underway in corporate affairs between functions elevated to genuine strategic roles and those quietly absorbed. The AI visibility dimension makes that divide significantly more consequential.

The organisations that reduced their communications capability did not just lose reputational resilience. They likely also lost the journalist relationships that generate earned media. The calls that stopped coming because there was no longer anyone worth calling. The coverage that did not happen. The authority that did not compound.

They are now paying the cost twice: once in the traditional reputational environment, and again in the AI-driven information environment, where their absence is being filled, often inaccurately, by whatever the AI can retrieve from whoever did maintain a presence.

Negative media coverage and mishandled crises now live permanently in AI databases, resurfacing without context or timeline. The organisations without the capability to manage their narrative proactively are not just invisible. They are vulnerable to being defined by what they failed to address.

The question worth asking now

Quantity versus quality has always been at the heart of the earned media measurement debate. Coverage breadth is the new KPI, not the number of placements, or the reach of an article, but how consistently an organisation surfaces across the AI ecosystem. That breadth is built through journalist relationships, compounding over years. It cannot be replicated by a content strategy, purchased through advertising, or reconstructed quickly once it is lost.

Leaders reading this have access to a simple diagnostic. Ask each of the major AI platforms who the credible voices are in your sector, who the leading organisations are, and who is shaping the conversation. Note what comes back and what doesn't.

If your organisation is absent, inaccurate, or thinner than your competitors, that is not a content problem. It is a structural one. And it will not solely be resolved by a campaign.

The journalists who would place you accurately in the AI ecosystem are out there. The question is whether your organisation has the relationships, the capability, and the function to reach them.