Erin Lanuti, the co-founder and chief executive of Lilypath, says that artificial intelligence is now acting as a filter for professional standing, judging a person’s knowledge and trustworthiness by pulling data from a variety of online sources. Consequently, an individual’s reputation can be formed before a researcher ever clicks on a corporate site, reads an executive’s biography, or opens a LinkedIn page.
Lanuti’s team has built a collection of tools under the banner Authority Intelligence that reveal how current AI models assess a professional’s authority. The system scans publicly available material—such as LinkedIn profiles—and delivers a full briefing that shows how AI interprets a person’s expertise.
How AI-Mediated Research Works
Lanuti explains that AI-mediated research focuses on the conclusions drawn after a model locates a person, not merely on the fact that the person is found. For those handling communications, this means monitoring not only traditional news coverage or Google rankings but also the way AI engines phrase and characterize their clients in generated answers.
Read Also: Instructor Uses AI to Replace Ten Professionals
A March 2026 study by the AI-search firm Profound identified LinkedIn as the most frequently referenced domain for professional queries on leading AI search services, illustrating the platform’s influence on AI’s perception of experts. Lilypath’s data consistently shows LinkedIn occupying the top spot, with YouTube close behind, across a sample of roughly 30 professionals and about 30,000 citations.
The dataset spans a wide range of occupations, including finance leaders, elected officials, physical-therapy clinicians, and consultants. Beyond the two primary sources, additional inputs such as corporate websites, reputable news articles, personal sites, by-lined pieces, professional bodies, conference bios, podcasts, transcripts, and trusted directories can all reinforce the narrative.
Authority Intelligence and Its Applications
Authority Intelligence measures and refines the way AI portrays an individual and their skill set. Picture an AI drafting a profile using clues it discovers online, LinkedIn listings, personal webpages, published articles, interviews, and more. At times the depiction is accurate; at other times it omits key details, conflates identities, or fails to mention the person entirely.
Lilypath examines that “report,” matches it against verified facts, and highlights adjustments that can improve accuracy, whether the AI’s answers evolve or stay static over time. The solution suite serves professionals at every career stage, with Blueprint programs ranging from college students to C-suite leaders.
Read Also: Halloween contact lenses get spooky makeover
Authority 360 targets single professionals, founders and executives. It pairs two annual LinkedIn Blueprint updates with ongoing Progress Checks and a weekly AI Pulse that monitors ChatGPT, Gemini and Claude. Authority Graph, the enterprise-level offering, evaluates ChatGPT, Claude, Gemini and Perplexity against a controlled Source of Truth, logs changes over time, uncovers the citations influencing each answer, and benchmarks individuals against comparable peers.
Lanuti notes that communicators are shifting from merely shaping messages to curating the evidence that algorithms draw upon. Traditional assets, media coverage, executive bios, authored content, speaking engagements and corporate sites, remain vital, yet teams must now verify that those inputs create a coherent, verifiable record.
Communications teams are advised to begin with a controlled audit: devise eight to ten realistic questions a journalist, potential client, investor, recruiter or partner might pose, then pose the same queries to ChatGPT, Gemini, Claude and Perplexity (enabling web search where possible). Review five key indicators: Is the subject present? Is the description correct? Does the answer acknowledge the true expertise? Is the person recommended or merely cited? Which rivals or sources appear instead?
Read Also: ARIZA Showcases Carbon Monoxide Alarms in Dubai
Traditional online reputation management usually asks: What appears when someone searches my name, and is the overall impression positive or negative? Authority Intelligence reframes the inquiry, asking instead: Does the model correctly identify the individual? Does it grasp the genuine skill set? Can it validate the claims? Does it suggest appropriate opportunities? Which sources underpin its conclusions?
The objective is to make a precise interpretation easy to retrieve and substantiate, rather than simply crafting a flattering impression. Publishing more content is not a guaranteed fix; information must be accessible, specific, consistent and corroborated elsewhere.
Lanuti’s research indicates that no provider can promise an AI will retrieve, employ, or cite a particular piece of material. The key insight is that there is no universal hierarchy of sources; relevance shifts with the model, the query, the freshness of data and the surrounding context. The strongest position, according to Lanuti, is a clear professional dossier supported repeatedly by multiple reputable sources.
