AI rewrites who gets seen, not just how stories spread

Earned media used to follow a clean path: pitch a journalist, land the story, watch it run. The value came from placement and the ripple through social channels or search. That model assumed human discovery at every step. Chatbots and AI-generated summaries now sit between the story and the reader. A great placement no longer guarantees an audience will find it the same way. The Reuters Institute brought together 17 experts to forecast how AI reshapes news by 2026. The short version: original reporting still matters, but the routes to readers are splintering fast.

LLM-generated summaries strip source credibility

Ask ChatGPT or Perplexity for news on a topic and you get a synthesized answer pulled from multiple outlets. The AI might cite sources in a footnote or inline link, but the framing and editorial choices belong to the model, not the publication. These tools are becoming a primary entry point for news discovery, especially among younger and time-pressed audiences. A brand mentioned in the Financial Times and a brand mentioned in a blog post can appear with equal weight inside an AI summary. Credibility becomes harder to signal when the original byline and publication context disappear. This flattens the earned media hierarchy. A tier-one placement still carries strategic value for journalist relationships and industry perception, but its consumer-facing impact is harder to predict. If the AI strips the publication's voice and repackages the facts, the brand's association with that outlet weakens in the reader's mind.

Fewer, deeper journalist relationships matter more now

Spray-and-pray pitching has always been inefficient. Now it's also becoming strategically irrelevant. If AI intermediates how audiences find stories, the publications that feed those models matter less than whether the brand shows up in the synthesis at all. PR strategies need to prioritize original, substantive reporting over clip volume. Journalists who break news, conduct interviews, and add analysis are more likely to get cited by AI models as authoritative sources. Brands that feed those stories with real insight, data, or access will show up in the AI-generated summaries that follow. This isn't about gaming the algorithm. AI rewards depth over distribution. A single investigative piece or deeply reported feature shapes how an LLM represents a topic more than a dozen newswire mentions. Brand relationships with journalists need to reflect that shift: fewer pitches, more exclusives, more time invested in helping a reporter build a better story.

Content consumption splinters across workflows, not platforms

Most PR measurement assumes audiences consume content on a platform or publication they chose to visit. AI breaks that assumption. Users increasingly encounter brand mentions inside tools they use for other purposes: productivity apps, search replacements, voice assistants. Experts see this trend accelerating, with AI-generated answers embedded in workflows rather than treated as a separate destination. A user might never click through to the original article but still absorb its key points through a summary generated on the fly. This makes attribution harder and earned media value murkier. A mention in a major outlet may never register as a visit, a share, or a conversion, even if it shapes perception. The old metrics like unique visitors, time on page, and referral traffic become less useful when AI reshapes and redistributes the content before it reaches the end user. Earned media doesn't lose value, but the evidence of that value becomes harder to collect and easier to misread.

Original reporting becomes the only moat

If AI-generated summaries dominate discovery, the publications and stories that get cited most often will be the ones that break news or add proprietary insight. Outlets that rewrite press releases or aggregate other reporting get squeezed out of the synthesis. The same logic applies to brand-owned content. A blog post that repackages industry trends won't make it into an LLM's answer. A data study, a proprietary survey, or a piece of cultural commentary that other outlets cite might. PR strategies need to adjust here. Brands that want to show up in AI-mediated news discovery need to act more like publishers: commission research, develop points of view, create the kind of material that journalists treat as source material. The goal isn't to replace earned media but to generate the raw material that makes earned media more substantive and more likely to be referenced when an AI assembles an answer. We've seen