Attribution, authority and AI

August 2025: PR has longed carried the burden to prove its worth. For years, the argument has been made that one intentionally placed article in a top tier media regularly consumed by your target audience is worth far more than a dozen low-value mentions. But when the inevitable question of “Show me the evidence” is asked, the answer has often been complicated. But that’s now changing.

With the rise of LLM search – the technology powering ChatGPT, Claude, Perplexity and others – the value of quality coverage is becoming clearer. These tools are now a daily habit for millions. OpenAI reported this month that it’s now operating at  700 million weekly active users, and according to Microsoft, 75% of knowledge workers now using generative AI as part of their work. Crucially, when these tools answer professional questions, they overwhelmingly lean on established, credible media sources.

That FT Big Read, the TechCrunch news piece, the Guardian interview, they no longer disappear once the news cycle moves on. They live on as part of AI’s knowledge layer, cited back in answers and shaping decisions long after publication. For the first time, we can see the attribution trail of how earned media informs real-world responses. This is tangible, boardroom-ready proof of PR’s long-term value.

A breakthrough, but the landscape continues to evolve

This shift is a breakthrough and has led to growing push in Generative Engine Optimisation (GEO). It validates what PR has been arguing for decades: that quality, trusted coverage carries significant influence, that can’t be explained via AVE. And it finally tips the scales in favour of quality over quantity.

But like everything, it’s not a finish line. The media world is still in flux. Legacy titles continue to restructure under commercial pressure. At the same time, person-led platforms are booming. Substack, valued at over $1 billion in July of this year, now counts 20 million monthly readers. This dual shift means the definition of authority is broadening, and PR has to evolve alongside it. For comms, this means:

• Be intentional with outlets. Prioritise placements in publications that carry authority with both human readers and AI models.
• Data in, visibility out. Make content data-backed and referenceable to increase the likelihood of it being surfaced by AI.
• Think longevity. A standout piece of coverage can influence over months, not just days.

Remaining focused on the target audience

While media coverage is a tangible output, it isn’t the end goal. The real measure is how it shapes the decisions people make. That’s where intention comes in. Reviews, customer case studies, analyst insights, proprietary data, these formats can hold strong influencer under AI scrutiny and become the reference points that can inform many.

As generative AI becomes embedded in daily working life, quality coverage is no longer reaching readers once, it’s feeding the very systems guiding business decisions. For PR, that means prioritising the right outlets, creating content designed to be referenceable, and thinking long-term.

By Will Cook, Founder