More than a third of consumers now begin product research with ChatGPT, Perplexity, or Gemini — not Google. Not a publication. Not a friend's recommendation. They type a question into an AI engine and take the answer seriously. That answer comes from somewhere. The brands and executives who shaped what got written, cited, and indexed are the ones who appear in it. The ones who didn't are invisible.
This is the structural shift that defines the next decade of communications work.
The Audience Is Now the Machine
I've said it before: AI Communications is a mix of journalism, psychology, and engineering — and the audience is now the machine. That's not a metaphor. It's a literal description of what PR has become.
When a buyer asks an AI engine "what's the best crisis communications firm" or "which beauty brands are leading on sustainability," the answer isn't pulled from a database of paid placements. It's synthesized from everything the model was trained on and everything it can retrieve — earned media, Wikipedia entries, research citations, trade coverage, analyst reports. The brands and people who dominate that corpus get cited. Everyone else doesn't exist.
This is Citation Share — your share of the answers AI engines now deliver to buyers in your category. It's the new market share metric for communications. And it's measurable.
Six Things AI Actually Changed in PR Work
I'll skip the thirteen-item listicle other people write. Six things matter.
1. First drafts collapse from days to minutes. Press releases, blog posts, bylines, talking points, pitch emails, social copy — first-draft time drops 30 to 50 percent. That is real. The savings reallocate senior time toward the work that earns the retainer: counsel, narrative, relationships.
2. Media targeting went from spreadsheet to signal. AI-powered platforms rank journalists by past coverage, sentiment alignment, and topic fit. The era of blasting 400-name lists is finally over for the firms that adapted. For the firms that didn't, it's a slow extinction event.
3. Real-time monitoring became a 24-minute discipline. The news cycle isn't 24 hours anymore. It's 24 minutes. AI sensing platforms catch emerging narratives before they break, and the firms that built that infrastructure win every crisis. The firms that didn't are still reacting to coverage that's already viral.
4. Measurement finally connects to business outcomes. Impressions and AVE are obsolete. AI-powered dashboards correlate earned media to web traffic spikes, branded search velocity, lead generation, and AI engine citations. PR is being measured the way digital advertising is measured — and the discipline is stronger for it.
5. A new core function showed up: AI visibility. Measuring and improving how brands appear inside ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, and Microsoft Copilot. Buyers research vendors inside answer engines before they ever talk to your sales team. The brand cited in those answers wins the consideration set. The brand absent loses, regardless of which product is better. This is the most important new PR capability of the decade.
6. The economics of agency work changed. The shift isn't headcount reduction. It's reallocation. Senior practitioners stop doing the work that AI does cheaper and faster, and start doing more of the strategic work the agency was always paid to do but rarely had time for. That's the real change.
Where Humans Still Win — and Always Will
AI drafts a press release in seconds. AI cannot tell you whether you should issue one.
The functions that still require human judgment are the same ones that have always defined senior counsel:
- Advising a CEO during a sensitive crisis.
- Deciding whether silence is the right move.
- Navigating product recalls, executive misconduct, regulatory pressure.
- Reading a journalist or a regulator or a board member in real time.
- Knowing when a story needs cultural nuance the model cannot fake.
- Building genuine relationships across journalists, analysts, regulators, customers.
These are not areas where AI will gradually catch up. They're the work of judgment, empathy, and accountability. Machines don't do those.
What This Means for How PR Gets Done
Traditional PR built awareness through placement volume. Get in The Wall Street Journal, get on CNBC, move the needle. That still matters — but it matters now partly because those placements feed the corpus that AI engines draw from. A Forbes feature isn't just a Forbes feature anymore. It's a retrieval anchor.
The implications are significant.
Earned media strategy has to account for AI retrievability. A placement in a high-authority publication that AI engines trust is worth more than ten placements in outlets they don't index heavily. Entity density, structured content, primary source citation — these are now PR strategy considerations, not just SEO ones.
Thought leadership has to be built for the machine first. Op-eds, bylines, research studies — the content that positions an executive or brand as the authoritative answer to a category question — need to be structured so AI engines can extract and cite them. That means clear claims, named entities, specific data, and consistent positioning language repeated across enough credible sources that the model recognizes it as consensus.
Wikipedia is back at the top of the priority list. It's one of the most heavily weighted sources in LLM training. A well-maintained, accurately sourced Wikipedia presence for a brand or executive is now a direct input into AI answer quality.
Crisis PR has a new front. Reputational damage now accumulates inside AI answers — and it's sticky. A brand that gets associated with a crisis in enough indexed content will find that association baked into AI responses long after the news cycle has moved on. Getting ahead of the AI narrative is as important as managing the press narrative. The standing framework: Crisis PR & Crisis Communications pillar.
The Biggest Mistakes Firms Are Making
The first: treating AI as a tool, not infrastructure. A tool gets bolted onto existing workflows. Infrastructure rebuilds them. The firms that rebuilt are pulling ahead. The firms that bolted on are about to learn the hard way.
The second: skipping governance. AI without an editorial review process eventually produces a hallucination that costs a client. AI without enterprise-grade data handling exposes client confidentiality. AI without disclosure rules erodes trust. The leading firms have written AI policies, mandatory human review on all external-facing content, enterprise deployments with data isolation, and clear disclosure standards. Every other firm is one bad output from a public mistake.
The third: ignoring AI visibility. If your firm tells clients how to use AI but doesn't measure or work on their Citation Share inside the AI engines, you're competing in last decade's discipline. The discipline that matters now is being the brand the engines cite.
Most PR firms are still optimizing for yesterday's distribution channels. They measure success in impressions, placements, and share of voice in trade publications. Those metrics aren't wrong — but they're incomplete. The firms doing serious work right now are measuring Citation Share across AI engines, tracking how often their clients appear in AI-generated answers to the category questions buyers are actually asking. The standing measurement framework is The EPR Citation Share Index.
That's where 5W AI Communications is focused. Generative Engine Optimization — GEO — is the discipline of building a brand's presence in AI-generated answers the same way SEO built presence in search results. It's technical, it's content-driven, and it requires understanding how AI engines weight sources, structure answers, and update over time. The full operating model is in The GEO Operating Stack.
The brands that invest in this now will have a structural advantage that compounds. The ones that wait will be trying to catch up against competitors who've already been cited hundreds of thousands of times in AI answers their buyers are reading.
The Fundamentals Don't Change. The Infrastructure Does.
None of this means the core of PR changes. Relationships with journalists still matter. A great story still travels. Crisis judgment is still mostly human. The instinct for what the public will accept and what it will reject — that doesn't get automated.
What changes is the infrastructure underneath. The distribution channels, the measurement frameworks, the content strategy, the technical layer that determines whether a brand's earned narrative actually reaches the AI engines that now sit between companies and their buyers.
The professional who uses AI replaces the professional who doesn't. The firm that uses AI as core operating infrastructure replaces the firm that treats it as an experiment. We're three years into the rebuild. The future arrived. The infrastructure exists. The discipline is named. The only question left is whether your agency is built for it.
