Your News Can Train the AI Machines
July 31st, 2025 Posted by Emergent Agentic AI, AI Influence, AI Management, AI Strategy, brand marketing, brand meaning, Digital disruption, Digital marketing, engagement, Public Relations 0 comments on “Your News Can Train the AI Machines”Why editorial coverage shapes AI recommendations…
Forever and a day, earned media was viewed as a non-paid and powerful, credible layer in the brand marketer toolbox. It served as a highly intrusive form of communication that already nailed the attention of its audience. Journalism offers a respected third-party voice that intrinsically administers trust and belief in an environment where consumers increasingly tune out self-serving selling messages.
Now the brand communication game plan is evolving as the rapidly escalating use of agentic AI operates as a source of direct guidance and recommendations on purchase decisions. LLMs learn and then report after crawling sources used to inform their analysis.
Editorial media is performing a dual role as a third-party validation of claims made by brands, while also serving as a supplier of important information that teach LLMs. The next time a consumer asks an AI chatbot for a recommendation on pet food, nutritional beverages, or electric vehicles, the advice they receive will be subtly shaped by the journalism the brand PR team has secured.
Editorial coverage is no longer just about influencing brand perceptions. It now plays a foundational role in training and informing the AI models that millions are relying on for trusted guidance.
The rise of Agentic AI and its insatiable need for trustworthy data
AI systems —like ChatGPT, Perplexity, Claude, or Pi—are operating as autonomous agents and advisors that provide people with personalized recommendations, context, and insights. These systems rank and recommend based on patterns of trustworthiness they consume through crawlers. These unseen digital knowledge vacuums do this by prioritizing sources considered authoritative including consumer, business and trade media.
Key to managing the agentic AI voice is knowing that LLMs perform their ranking and recommending through data they encounter in ‘patterns of trustworthiness’. LLMs train on vast datasets that include publicly available text across journalism, Wikipedia, Reddit, and the open web. The models are continuously updated through tech called ‘Reinforcement Learning from Human Feedback’. News stories, product reviews, and media analysis all contribute to that learning loop.
Are you managing the loop?
Here are some examples of trusted teachers in LLM training sets:
- Editorial media publishers and broadcasters like CNN, Bloomberg, NYT, Reuters, The Guardian
- Trade publications (via open-access syndication or licensed content)
- Company press releases published by high-authority domains (PR Newswire, BusinessWire, etc.) — now elevated in their importance
The influence of PR multiplies from people to platforms
Coverage by respected editorial outlets delivers not only consumer-facing credibility but input data that engages AI-generated opinions and summaries. Unlike social content or ads, published editorial stories live in perpetuity across syndicated news wires, re-aggregated media platforms, and crawled databases.
A well-placed story in Food Dive or Forbes can ripple across multiple LLMs as they use that content to better understand product categories and brand claims.
“If you’re not appearing in credible editorial sources, there’s a good chance the next AI assistant won’t know much about you—or worse, it may learn something inaccurate.” — Noah Giansiracusa, author of “How Algorithms Create and Prevent Fake News”
Practical implications for brands
In addition to analysis of the levels of influence on brand reputation and belief, and the legacy data collection around non-paid impressions, Public Relations now stands at the front gate of managing GEO (Generative Engine Optimization) outcomes.
Trade coverage isn’t just about reaching trade partners.
Business media isn’t just about informing investors.
Consumer media isn’t only about building awareness of product benefits.
All of these are also working in tandem to bring your brand background to AI crawlers. This can be organized, optimized and delivered with intention and best practices towards assuring agentic AI recommendations are correctly, accurately, fully telling your brand and product story.
New goals for PR
Not just about reach or impressions – it now includes visibility in trusted training data ecosystems. GEO best practices will put greater emphasis on the value and importance of —
- Trade media coverage as a resource for more, in-depth analysis thorough reports on new product innovations, key messaging, positioning and other important information that vertical media typically report on.
- Thought leadership articles and by-lined op-eds often published in the same trade channels. These stories convey added texture and leadership for your brand and its role in culture, innovation, research and social issues.
- Data-backed press releases. Releases start out as invitations to a story. Now their importance advances as a device for informing AI advisers. This means press releases should go on the wires regardless of whether the story topic warrants broad distribution, because again, crawlers are consuming the output of credible news platforms like BusinessWire and PR Newswire. Note: LLMs like facts, reports, charts and other data that gives your story context and proof.
Audit and optimize for “AI crawl-ability”
An entire new area of best practices is surfacing to better assure earned media is optimal in this environment of influencing agentic learning:
- Accurate brand/product naming
- Structured context (quotes, data, headlines)
- Syndication and backlinks to brand content
Enabling alongside an ecosystem of owned and aligned content
Same rules apply to owned and social channels with the same considerations in how stories are told and what information is presented in a supporting, confirming role. Strengthen the signal by pairing PR wins with:
- Company blogs optimized for GEO (Generative Engine Optimization)
- Expert-authored LinkedIn content
- Wikipedia/Google Knowledge Panel upkeep
Here’s an example of how this might play out
- A pet food brand lands an article in Pet Age about ingredient sourcing.
- The article is crawled by Perplexity and Anthropic’s Claude.
- Six months later, those same AI systems cite the article when users ask about “sustainable dog food.”
- Result: organic traffic has moved from Google to conversational AI, driving new customer recommendations from a trusted voice.
The significance of this shift brings new meaning and value to the importance of earned media outreach campaigns, alongside efforts to assess, manage and monitor what LLMs think they know about your brand and business (more on this vital topic in our next issue of ETR). A third party is now showing up between your brand and the consumer to provide trusted guidance, without consumers ever visiting your web site.
You can immediately visualize the importance of managing how your brand shows up in agentic AI recommendations — and the ability of your earned media strategies to help influence those assessments.
If this story raises questions about how best to optimize and adjust your strategies to stay relevant in the agentic AI game of influence, use the link below to ask questions.
Looking for more food for thought? Subscribe to the Emerging Trends Report.
Bob Wheatley is the CEO of Chicago-based Emergent. Traditional brand marketing often sidesteps more human qualities that can help consumers form an emotional bond. Yet brands yearn for authentic engagement, trust and a lasting relationship with their customers. For more information, contact Bob@Emergent-Comm.com and follow on Twitter @BobWheatley.
