In a recent study released by Pangram Labs, journalists across the United States expressed a clear preference for human‑authored sources when responding to media‑relations requests. The data, gathered from Featured’s three platforms—Featured, HARO, and Connectively—covers the period from June 2025 through July 2026.
Scale of the problem
During that 13‑month window, the platforms received a total of 1,394,740 pitches in response to 42,392 journalist inquiries. Pangram’s AI‑detection tool flagged 292,895 of those pitches, roughly 21%, as fully AI‑generated.
Journalist response to AI‑generated content
Among more than 2,500 journalists surveyed, 35% actively chose an option to block AI‑generated pitches, even though the default setting allows such pitches to appear. When journalists did not opt out, the selection rate for AI‑generated pitches remained low: only one in ten AI pitches were chosen for publication, compared with about one in three human‑written pitches.
Impact on source accounts
The study also highlighted a correlation between AI‑generated content and account suspensions. Banned accounts were 3.6 times more likely to have submitted an AI‑flagged pitch than active accounts—24.7% of banned accounts had at least one fully AI‑generated pitch versus 6.8% of active accounts.
Platform safeguards
While HARO does not automatically ban accounts for submitting AI‑generated pitches, the presence of such content can trigger internal monitoring. The platform filters out accounts that repeatedly submit large volumes of low‑quality AI responses in an effort to preserve trust among journalists.
Why some turn to AI
Companies and individuals seeking earned media exposure sometimes employ AI‑driven public‑relations services to mass‑produce responses to journalist requests. The goal is to increase the likelihood of landing press features without the time investment required for genuine expert outreach. However, the data indicates that this strategy rarely succeeds on Featured’s platforms.
Notable example
The report cites the case of Dr. Eleni Nicolaou, an alleged art therapist whose fully AI‑generated quotes appeared in more than 30 publications, including Vice, Forbes, and Glamour. Investigations suggest the persona may not exist at all, underscoring the risk of misinformation.
Conclusion
Overall, the findings suggest that AI‑generated PR pitches are not meeting the quality bar expected by journalists. Human expertise and authentic voices continue to dominate earned media coverage, and platforms are taking steps to ensure that AI‑generated content does not undermine the credibility of the source‑matching process.
Original reporting: KRDO (Colorado Springs metro) — read the source article.
