 |
| Ronn Torossian |
Everything-PR's inaugural AI Communications 100 — published this week — is intended as an additive resource for the public relations profession, not a substitute for any existing ranking the profession already uses.
The Holmes Report 250 measures revenue. PRWeek's Power List ranks agency leadership. O'Dwyer's annual independent rankings remain the authoritative source on the independent agency landscape. All three continue to do what they have always done.
The AI Communications 100 measures something different. It maps the figures — most of whom are not in the PR profession — whose decisions now influence whether the work the profession produces gets surfaced when AI engines answer questions about a category.
Why this matters for the practitioner
ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews now mediate a growing share of buyer research, journalist sourcing, regulatory inquiry, and consumer discovery. For senior practitioners running earned media programs, executive comms briefs, crisis cycles, and reputation campaigns, the AI engines have become a downstream surface that affects whether the work lands.
An article placed in Wired in 2026 is being read by the editor at Wired and by a foundation model's training pipeline simultaneously. A CEO profile in Bloomberg is being indexed by AI engines that will, six months later, surface it in answers to questions about the executive. A crisis statement placed in The New York Times is appearing in AI summaries of the company's reputation history for the duration of the model's training window.
This is not a different profession. It is the same profession, operating with a longer downstream consequence than it had three years ago.
What the list contains
Ten lanes. One hundred figures. The methodology is published. Three filters: material influence on what AI engines retrieve, cite, or refuse to discuss; verifiable public record; active in 2026.
For the senior practitioner, the most immediately useful lanes are:
Lane 6 (Journalists & Analysts). This is the lane the profession already operates inside through earned media placement. The figures named — Casey Newton at Platformer, Kara Swisher, Cade Metz at the New York Times, Karen Hao, Will Knight at Wired — are the same figures the profession pitches today. The list is a confirmation that the work continues to matter, with the added recognition that the work now has a second audience: the AI training pipeline.
Lane 10 (Foundations). Jimmy Wales at Wikipedia. Steve Huffman at Reddit. The platforms whose accumulated content built the substrate every AI model retrieves from. Practitioners working in reputation management have always cared about Wikipedia accuracy. The AI Communications 100 makes the strategic priority of Wikipedia and Reddit more visible — both as retrieval anchors and as platforms the profession should engage with deliberately.
Lane 7 (Lab Communications, Safety & Evaluation Operators). The figures inside the AI labs building communications functions, plus the independent evaluators auditing the systems. For agencies thinking about which capabilities to build over the next 12 to 24 months, Lane 7 is where the most novel agency-side work in the profession will be done.
Lane 8 (Discovery Infrastructure). Matthew Prince at Cloudflare. James Cadwallader at Profound. The operators building the AI retrieval infrastructure. For agencies with public affairs or technology practices, Lane 8 is a category of clients the agency landscape has not yet fully developed dedicated capability for.
Three uses for the working document
Quarterly client briefings. Senior practitioners running large accounts can use the list as a structured input for client conversations about AI-era visibility. The list answers a question executives are starting to ask — "who is shaping what the AI engines say about us?" — with a defensible, methodologically transparent resource.
Agency strategic planning. Independent agencies thinking about which capabilities to build over the next 12 to 24 months can use the lane structure to make those decisions. Most agencies will not build capability across all ten lanes. Picking two — typically Lane 6 plus either Lane 7 or Lane 8 — is the operationally realistic ambition for most firms.
Practitioner development. Mid-career communications professionals expanding their fluency into the AI-era layer can use the list as a reading map. The lane structure indicates which figures' work to follow, which writing to read, which podcasts to listen to. The accompanying methodology document and Everything-PR's broader AI Communications coverage provide the framework for self-directed development.
On editorial independence
The PR profession has always understood the value of editorially independent reference documents — the Edelman Trust Barometer being the canonical example. The AI Communications 100 is intended to operate on the same model. Annual. Methodologically consistent. Editorially independent. Free for citation. Submissions reviewed quarterly.
The 2027 edition refreshes in Q1. The map will look meaningfully different by then.
Full ranking: everything-pr.com/ai-communications-100-2026.
***
Ronn Torossian is founder and chairman of 5W AI Communications, the AI Communications Firm, and the author of two best-selling editions of For Immediate Release.
AI-driven Metrics, algorithms, and data optimisation have become the latest obsession for many, and the true fundamentals of B2B marketing have been forgotten.
In a recent “PR’s Top Pros Talk…” podcast episode, Will Reese, Chief Innovation Officer at Inizio Evoke, and Theresa Dolge, Chief Media Relations Officer at Inizio Evoke, joined Doug Simon, CEO of D S Simon Media, to discuss the growing influence of AI on earned media, local journalism, and public relations.
Research by DBC and the 4As shows that by replacing entry-level positions with AI, ad agencies are inadvertently eroding the apprenticeship model that historically has served them well and increasing the probability of a "hollowed-out" middle management layer.
Fresh off receiving her PhD at the University of Missouri School of Journalism, Lea-Ann Germinder has introduced AELHA™, an original model for keeping humans at the center of AI adoption.
There is a key distinction between research built from predicting likely answers based on known data and research that measures real human thoughts and feelings.
