The online conversation about your organization — every byline, public thread, news article, video — is teaching AI how to talk about your brand.
ChatGPT, Claude, and other large language models (LLMs) are trained on billions of web pages, including news articles, industry publications and other authoritative content (including social media platforms). If your company has a strong media presence, you’re already in their training data. Otherwise, you’re invisible in what may be the next generation of search.
Has the search game changed?
Traditional SEO taught us to optimize for Google’s algorithm so that we could show as a top search result. But AI doesn’t work like search engines. You can’t steer it with keywords or backlinks. Instead, LLMs learn about businesses the way humans do — through repeated exposure to quality information from trusted sources. And AI provides a specific answer rather than a list of links to click.
When someone asks an AI “What’s the best PR firm in Des Moines?” or “Which doctor is most trusted for my procedure?”, the answer comes from patterns in the model’s training data. Companies with consistent media in authoritative sources are more likely to be mentioned. Companies without that presence simply don’t exist in AI’s knowledge base.
Public relations can build immediate brand awareness and long-term knowledge infrastructure for AI.
What actually influences AI models
Not all media is equal. A mention in The New York Times, TechCrunch or an industry-leading publication carries significantly more weight than this blog post. LLMs are trained to recognize authoritative sources, and they retain information from those sources more reliably.
You can aim to answer specific questions with content you produce. You can implement structured data. Still, trust and authority matter; when your expertise gets published independently (conference videos, news quotes, etc.), that information gets encoded into how AI understands your company’s positioning and capabilities. This is the core of nascent GEO (Generative Engine Optimization) work pursued by PR teams and tech vendors alike.
The narrative built through public relations strategy shapes how AI contextualizes your brand. If every article about your company mentions innovation and leadership, that becomes part of your AI-mediated reputation.
Does social media content train AI?
Right now, social media content mostly shows up indirectly, mainly when referenced in trusted sources. And LLMs can’t see private profiles and communities. So, your executive’s Facebook post shows up when it’s referenced in an article, but not if it was published only to their friends. We know ChatGPT is crawling Reddit, LinkedIn, Substack and more. You won’t find real time access to this content, but LLMs are entering the data. Amid the unknowns, content that builds authority is the imperative.
A good catch: you can’t fake it
AI can’t be manipulated like search results. You can’t pay for better AI placement.
What works is genuine, sustained PR that creates real value — actual news, authentic expertise, and content that publishers want to share because it matters to their audience. The companies that will dominate AI-mediated discovery are the ones already doing PR right: building relationships, creating newsworthy moments, and establishing true leadership.
Bottom Line
Your public relations strategy should earn attention and trust today while building a foundation of knowledge that will determine whether AI includes you in queries happening over time.
Organizations that recognize this shift now and invest in a consistent, authoritative media presence will have an enormous advantage as AI increasingly guides discovery and evaluation. And the rewards of building your reputation will be realized even if an AI-first world doesn’t materialize.