The public relations profession has been quietly expanding for the last 36 months. Most practitioners have been so focused on running their actual work — client deliverables, crisis cycles, executive comms briefs, agency growth — that the expansion has happened around them rather than to them. The [AI Communications 100](https://everything-pr.com/ai-communications-100-2026) — Everything-PR's inaugural annual ranking of the figures shaping what ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews retrieve and answer — is one attempt to make the expansion legible. The list does not replace the existing power lists in the profession. The Holmes Report 250 still measures revenue. PRWeek's Power List still ranks agency leadership. The Arthur W. Page Society still names the chiefs at the Fortune 500. All three remain accurate measurements of what they measure. What the AI Communications 100 measures is different. It is a map of the operators — most of them outside the PR profession itself — whose decisions now affect whether the work the PR profession produces gets surfaced when buyers, journalists, regulators, and customers ask AI engines about a category. This is not a critique of the existing profession. It is a recognition that the profession's surface area has expanded. ## What the expansion looks like For most of public relations history, the categories of communications work were stable: media relations, executive comms, internal comms, crisis comms, public affairs, investor relations, reputation management. Those categories remain. New categories have emerged alongside them — and in some cases inside them. AI Communications. Generative Engine Optimization. Citation Share measurement. Lab communications. Discovery infrastructure communications. Each new category sits at an intersection that did not exist when the profession's organizational chart was last redrawn. AI Communications sits between PR and engineering. [GEO](https://everything-pr.com/what-is-generative-engine-optimization-the-complete-2026-guide) sits between PR and SEO. [Citation Share measurement](https://everything-pr.com/citation-share-the-new-discoverability-kpi) sits between PR and analytics. Lab communications sits between PR, product development, and AI safety research. The figures on the AI Communications 100 are the people doing the work in these new intersections. The traditional power lists do not name them not because the lists are wrong but because the figures are operating in categories the lists were not built to measure. ## Why the list is structured the way it is Ten lanes. One hundred figures. The lane structure is intentional — and it matters more than the rankings within lanes. Lane 1 (Lab & Infrastructure Principals) names the figures whose product and strategic decisions shape what the models can do at all. [Sam Altman](https://openai.com) at OpenAI. [Dario Amodei](https://www.anthropic.com) at Anthropic. [Demis Hassabis](https://deepmind.google) at Google DeepMind. [Jensen Huang](https://www.nvidia.com) at Nvidia. These are not communications professionals. They are the upstream decision-makers whose work makes everything in Lanes 2 through 10 possible. Lane 2 (Answer Engine Builders) names the figures shaping how the model output reaches the user. [Aravind Srinivas](https://www.perplexity.ai) at Perplexity. Liz Reid at Google Search. Mustafa Suleyman at Microsoft AI. The decisions made in Lane 2 about citation behavior, source treatment, and response format directly affect what the user reads. Lane 6 (Journalists & Analysts) names the reporters whose coverage becomes part of the training data the models retrieve from. This is the lane the PR profession already operates inside — through earned media placement — but the downstream consequence of that work has changed. An article placed in The New York Times in 2026 is being read by the editor at the Times and by a foundation model's training pipeline simultaneously. Lane 7 (Lab Communications, Safety & Evaluation Operators) names the figures doing the most novel communications work in the industry. [Jan Leike](https://www.anthropic.com) at Anthropic. [Beth Barnes](https://metr.org) at METR. Hannah Wong at OpenAI. The audience for their work is not a journalist or an investor. It is the model itself, and the humans who will subsequently ask the model about the company that built it. This is a category of work that did not exist five years ago. Lane 8 (AI Discovery & Visibility Infrastructure) names the figures building the pipes through which content flows into AI answers. [Matthew Prince](https://www.cloudflare.com) at Cloudflare. James Cadwallader at [Profound](https://profound.ai). [Edo Liberty](https://www.pinecone.io) at Pinecone. [Harrison Chase](https://www.langchain.com) at LangChain. The communications profession has not previously had to engage with infrastructure operators of this kind. It does now. ## The methodology, restated Three filters govern inclusion. Material influence on what AI engines retrieve, cite, or refuse to discuss. Verifiable public record. Active in 2026. Curation by the Everything-PR editorial team. Submissions reviewed quarterly. Everything-PR does not accept payment for inclusion. The full [methodology](https://everything-pr.com/ai-communications-100-methodology) is published. The independence of the list is not a marketing claim. It is a structural requirement. [Edelman's Trust Barometer](https://www.edelman.com/trust-barometer) works because nobody can buy a higher score. The AI Communications 100 operates on the same principle. The credibility of any ranked editorial product depends on it. ## Three things for senior practitioners The communications professionals reading this list and recognizing that most of it operates outside the profession's immediate field of vision should treat that recognition as data, not as criticism. Three actions follow. **First, expand the field of vision.** The figures in Lanes 7 and 8 are operating in adjacent categories that the next generation of communications practitioners will need to be fluent in. The fluency is not technical. It is operational: understanding what crawler controls do, what vector databases are for, what GEO measures, how lab communications functions differ from corporate communications functions. The reading is available. Everything-PR publishes the [GEO Operating Stack](https://everything-pr.com/geo-operating-stack), the [AI Communications team playbook](https://everything-pr.com/ai-communications-team-playbook-90-days), the engine-by-engine guides, the methodology behind the Citation Share Index. The material is dense but not opaque. **Second, build cross-lane relationships.** The most consequential conversations in the profession over the next decade will happen between practitioners in traditional PR roles and operators in Lanes 7, 8, and 10. The agencies and in-house teams that build cross-lane relationships early will have the institutional knowledge to operate in the system. The teams that do not will be operating inside it without recognizing it. **Third, treat AI Communications as additive, not replacement.** The traditional disciplines — media relations, crisis communications, executive comms, internal comms — are not being replaced by AI Communications. They are being supplemented by it. A senior practitioner running a crisis cycle in 2026 still does crisis communications. That same practitioner now also needs to know what the AI engines have been saying about the crisis, who is shaping that conversation, and how the retrieval graph compounds over the cycle. The traditional discipline remains. The new layer sits on top of it. ## The list is a working document The AI Communications 100 is intended to be a working document, not a static reference. The 2027 edition refreshes in Q1. Submissions reviewed quarterly. The lanes themselves may evolve. Some current figures will scale into category leaders. Some will move on. New figures will emerge — particularly in Lane 7, which is expected to grow as the AI labs build proper communications functions, and Lane 3, which will grow as AI policy shifts from drafting to enforcement. The point of the list is not to be definitive in any single year. The point is to be a standing reference document that gives the communications profession a map of a system that did not exist five years ago and that is now consequential for every brand the profession serves. The expansion of the field is good news. It means the communications profession now operates in more places, with more consequence, with a larger surface area than it had even three years ago. The AI Communications 100 is one attempt to map that. The work the profession does still matters. It now also matters in more places. Full ranking and methodology: [everything-pr.com/ai-communications-100-2026](https://everything-pr.com/ai-communications-100-2026). --- _Ronn Torossian is the founder and chairman of 5W AI Communications, the AI Communications Firm. He is the publisher of Everything-PR and the author of two best-selling editions of For Immediate Release._