What Counts as a Credible Expert Voice

The research categorized sources within articles, distinguishing named individuals with identifiable credentials from anonymous corporate spokespeople or unattributed claims. Government officials, academic researchers, licensed practitioners, and recognized industry experts all served as credible voices. Unnamed executives or generic company representatives did not.

The distinction matters for how PR teams approach media placements. An article quoting a named medical researcher alongside a brand spokesperson carries different citation potential than an article quoting only ‘a company spokesman said.’ Both articles might appear in the same outlet. The sourcing structure determines which AI engines find more usable.

Retrieval systems are built to surface information that audiences can verify and trust. Named sources with identifiable expertise are easier to verify than anonymous corporate statements. Articles that include them are structurally more credible in ways that AI extraction logic can act on.

The Government Source Finding

Government sources appeared separately in the Citation Consistency model, with a beta coefficient of .109 (p < .01), making it the strongest predictor of consistent citation reuse in that model. Articles that included named government voices were cited repeatedly across different prompts and personas, not just once.

This reflects something PR teams already know about government sources in traditional media relations: regulatory agencies, public health authorities, and government officials carry institutional authority that journalists and readers trust. AI retrieval systems appear to encode that trust signal as well.

For sectors where government communication is frequent and relevant, such as pharmaceuticals, healthcare, defense, and public affairs, this finding reinforces an existing media relations instinct. Getting coverage that includes named government voices, or generating coverage that your brand appears alongside authoritative institutional sources, is associated with more durable AI citation behavior.

Why Corporate Commentary Alone Falls Short

The research explicitly notes that generic corporate commentary does not create the same citation signal as credible expert voices. A large share of earned media content is built primarily around brand spokesperson quotes, which makes this finding worth sitting with.

A company executive statement, however well-crafted, does not carry the same retrieval weight as a named researcher, physician, or regulatory official making the same claim. The brand perspective matters for traditional PR measurement: share of voice, message pull-through, spokesperson prominence in coverage. It matters less for AI citation behavior.

This creates a practical question for how PR teams structure media pitches. If the goal includes AI citation durability, the spokesperson strategy needs to consider whether the brand is the primary voice or whether third-party credible voices are built into the story architecture from the start.

Building Credible Voice Strategy into Media Relations

Three approaches help incorporate credible expert voices into earned media pitches in ways that serve both traditional and AI citation goals.

First, develop relationships with third-party validators before you need them. Researchers, academic partners, industry practitioners, and government contacts who are willing to be quoted in media coverage are more valuable now than they were when AI citation was not a factor. Build that network deliberately.

Second, include named expert voices in your pitch materials. When you provide a journalist with the story and also make credible third-party contacts available for interviews, you increase the probability that the resulting coverage includes the sourcing structure associated with higher AI citation rates.

Third, review your spokesperson strategy for earned media with AI citation in mind. Company executives remain important voices for traditional PR outcomes. For AI citation durability, consider whether each earned media opportunity also requires a named external expert, researcher, or practitioner to anchor the coverage’s credibility.

The Pharmaceutical and Healthcare Angle

The research focused on health and GLP-1 content, making the findings particularly applicable to pharmaceutical and healthcare communications. In those sectors, named physicians, clinical researchers, FDA officials, and patient advocates are already standard sources for strong earned media.

The data suggests that sourcing structure is doing double duty: it satisfies editorial credibility requirements and signals to AI retrieval systems that the article is worth citing. For pharma PR teams thinking about how AI search shapes what patients and healthcare professionals find about their therapeutic areas, this is a concrete, measurable lever.

Earned media that includes named KOLs, academic researchers from credible institutions, and references to government guidance is the kind of coverage that AI engines seem to return when audiences ask questions about those therapeutic areas. For teams working in this space, Fullintel’s pharmaceutical media monitoring capabilities provide the coverage intelligence needed to track these patterns.

Measuring the Expert Voice Signal

Tracking which of your earned media placements include credible expert voices, and which of those placements are generating AI citations, requires coverage analysis at a level of granularity most teams are not yet applying.

Fullintel’s strategic media analysis team can help communications leaders build that picture systematically rather than anecdotally. The goal is to understand your current coverage portfolio against the characteristics associated with AI citation, identify gaps in expert voice inclusion, and adjust your media relations strategy accordingly. The research gives you the variables. The measurement layer gives you the feedback loop.