Copyright Is the Headline; Capability Is the Blind Spot: AI Technology in the Book-Publishing Trade Press, November 2025--August 2026

Fred Zimmerman

REVIEW 3 major objections 4 minor 16 references

Copyright Is the Headline; Capability Is the Blind Spot: AI Technology in the Book-Publishing Trade Press, November 2025--August 2026

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The book-publishing trade press covers AI as lawsuits and launches, not as engineered systems, leaving publishers unable to evaluate what the technology can actually do.

desk verdict A useful, honest map of how the publishing trade press covers AI, but the headline 'capability blind spot' count is built on a generic depth scale that only partially measures the claimed gap. read the letter →

arxiv 2608.00964 v1 pith:GQEUINZ6 submitted 2026-08-02 cs.CY

Fred Zimmerman This is my paper

classification cs.CY

keywords AIinpublishingtradepresscoveragebooktechnicaldepthcapabilityevaluationjournalismgapretrieval-augmentedgenerationprovenance

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This rapid evidence review of 89 AI-and-publishing articles across eight languages from November 2025 through August 2026 finds the trade press is neither silent nor uniformly hostile—about 30% risk-framed, 42% mixed, 28% opportunity-framed—but it clusters on rights, licensing, governance, reader trust, workflow adoption, and product announcements. Only ten items offer sustained technical scrutiny, and no sampled article centers a direct interview with a frontier-lab researcher or evaluation engineer. The paper argues the missing beat is technical adversarialism: reporting that connects model architecture, evaluation, retrieval, agent reliability, inference economics, model drift, and provenance to publishing decisions. If correct, publishing decision-makers are making legal and commercial choices about systems whose capabilities and failure modes the press is not equipping them to assess. The review proposes a standing AI beat, recurring lab interviews, a claims ledger, shared test protocols, reader panels, and technical columnists.

What carries the argument

The coding instrument is a 0–3 technical-depth scale plus a dominant-voice category applied to each of the 89 corpus items; it converts qualitative impressions of shallow coverage into countable structure—mean depth 1.63, researchers dominating five articles, zero frontier-lab interviews. The seven missing beats then define what depth-3 reporting would look like: naming the full tested system, predeclared pass thresholds with human baselines, retrieval-attack coverage, delegation-aware workflow reporting, per-unit cost denominators, factorial reader experiments, and signed-manifest provenance.

What would settle it

Find, within the same November 2025–August 2026 window and the same source strata, a trade-press article that centers a direct interview with an OpenAI, Anthropic, or Google DeepMind researcher or evaluation engineer and sustains depth-3 technical scrutiny. Alternatively, code a probability sample of 200 or more articles from the same outlets and show that depth-3 prevalence is substantially above the paper's observed 10-of-89 rate.

Watch

Extended reading notes

Core claim

The paper's central claim is that the trade press has graduated from asking whether AI matters but still reports AI mainly as lawsuits, scandals, policies, and launches; it rarely establishes what a system can do, under which conditions, at what cost, with which failure modes, or for how long the answer remains valid. The load-bearing observation is the capability blind spot: only ten of 89 coded items reach depth 3 on the paper's 0–3 technical-depth scale, commentators average 2.14 versus 1.54 for trade reporting, researchers dominate only five articles, and no article is structured around an interview with a researcher or evaluation engineer from a frontier lab. The paper specifies seven m

Load-bearing premise

The field-level claim rests on the assumption that the purposive, single-coded sample of 89 items fairly represents the global trade press; the author explicitly states that it is an analytic sample, not a census, and that no intercoder reliability is claimed.

Editorial extensions

If this is right

If the blind-spot claim holds, current trade coverage leaves publishers poorly equipped for the agentic-AI step change that began in late 2025.
The recommended remedies—standing AI beat, recurring lab interviews, claims ledger, test kitchen, reader panel, and incident-reporting norms—would give publishing a technical accountability layer analogous to its existing financial or legal reporting.
The Chinese-versus-Anglophone comparison implies two one-sided frames: operational reporting can mute conflict and labor, while conflict reporting can miss infrastructure and implementation.
The seven missing beats function as a concrete reporting checklist that trade editors could adopt immediately without new technology.
Cases like Shy Girl and Daggermouth show that detector-based provenance is insufficient; upstream chain-of-custody and due process are the operational response the paper advocates.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A testable extension is to apply the same coding scheme to another trade press—music, film, or legal publishing—and compare depth distributions; the paper's claim predicts similarly shallow capability coverage there.
The 'claims ledger' and 'test kitchen' proposals could generalize beyond journalism into procurement: publishers subscribing to AI tools could demand the same baselines in RFPs and vendor contracts.
If the Chinese operational frame is as distinct as coded, cross-national editorial exchanges might import infrastructure reporting into Anglophone coverage and adversarial due-diligence into Chinese coverage.
The paper's own single-coder, purposive-sample limitation points to a preregistered multilingual content analysis as the direct next study.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know this paper for one reason: it's the first coded multilingual look at how the book-publishing trade press covers AI, and it backs its observations with a real corpus. The author found what any industry watcher would suspect: coverage clusters around lawsuits, licensing, scandals, and product launches, while technical capability, evaluation, and failure modes are thin. That finding is useful, and the paper's seven recommendations for editors are concrete and sensible. I also give it credit for being unusually honest about its own limitations: single coder, purposive sample, paywall interference, no intercoder reliability, and a supplement that is promised but wasn't inspectable. Not many papers say that clearly.

The soft spot is more specific than the usual sampling worries. The 'only ten items offer sustained technical scrutiny' number is the paper's main quantitative evidence for the 'capability blind spot.' But the depth-3 code was defined as scrutiny of mechanism, evidence, boundary conditions, and failure modes—a generic analytical depth standard. That can be satisfied by a careful legal explainer, which the paper itself praises, without ever touching model architecture, evals, RAG, prompt injection, or the specific missing beats the paper lists. So the 'ten items' count is not a clean measure of the capability gap. The separate claim that no article centers a frontier-lab interview is more direct and more robust. The qualitative reading of the articles and the multilingual comparison—particularly the Chinese operational frame vs. the Anglophone harm frame—do most of the real work. The count should be re-framed as a proxy or supplemented with a dedicated capability-code in a revision.

Is it worth a referee? Yes. It's a well-scoped, honest rapid evidence review with a new dataset and actionable conclusions. The construct-validity issue and single-coder reliability are addressable in revision, not fatal. For publishing professionals and anybody studying AI media coverage, this is a genuinely useful reference. For a CS audience, it's a case study, not a technical advance, but it's a careful one. I'd send it to peer review and ask for an intercoder check on a subset, a more systematic disclosure of the search protocol, and either a narrower claim or a new code that actually measures capability coverage. The paper's own recommendations—claims ledger, test kitchen, lab interviews—are good advice for the trade press, and the author would do well to follow the same standard.

Referee Report

3 major / 4 minor

Summary. This rapid evidence review examines 89 articles about AI and book publishing published between November 2025 and August 2026, drawn from a purposive multilingual corpus. The author codes each item for topic, stance, technical depth, and dominant voice. The paper reports that the trade press is neither silent nor uniformly hostile: 30% of items are risk-framed, 42% mixed, and 28% opportunity-framed. It also finds that coverage clusters around rights, licensing, governance, reader trust, workflow adoption, and product announcements; that only ten items reach the highest technical-depth score; that no item centers on a direct interview with a frontier-lab researcher or evaluation engineer; and that Chinese coverage is markedly more operational and opportunity-oriented. The author argues that the central gap is 'technical accountability': reporting that connects model architecture and evaluation to publishing decisions. The paper closes with concrete recommendations for trade editors, including a standing AI beat, recurring lab interviews, a claims ledger, shared test protocols, and reader panels.

Significance. If the central finding holds, the paper makes a useful and actionable contribution: it identifies a concrete, potentially consequential deficit in trade coverage of AI in publishing, and its recommendations are specific and practical. The paper's strengths include its multilingual corpus, the explicit disclosure of its purposive sampling and single-coder limitations, and the promise of a full coded corpus, claim ledger, and glossary in the supplement, which would allow independent audit. The 'no frontier-lab interview' claim is a falsifiable, interesting observation. However, the main quantitative support for the 'capability blind spot'—the depth-3 count—is construct-mismeasured, so the headline statistic does not directly support the paper's central claim as currently written.

major comments (3)

[Section 2, Figure 2, Abstract] The depth-3 code is defined as 'mechanism, evidence, boundary conditions, and failure modes receive scrutiny.' This is a generic intellectual-depth standard, not a capability/evaluation standard. An article that carefully explains a legal or process mechanism—e.g., the fair-use versus pirated-copies distinction praised in §3.1—can earn a depth-3 score without engaging model architecture, evals, RAG, prompt injection, agent reliability, or inference economics, i.e., the very beats the paper calls the 'central gap.' Thus the abstract's 'Only ten items offer sustained technical scrutiny' does not measure the 'capability blind spot' as defined. The statistic is construct-misleading. Please add a separate code for 'capability/evaluation engagement' or re-analyze the ten depth-3 items and show explicitly how many engage each of the seven missing beats.
[Section 4.1] The claims that the corpus lacks eval reporting, RAG attack-surface coverage, agent-risk distinction, unit economics, and factorial reader research are asserted without systematic counts or item-level evidence. For a gap analysis, absence claims require more than impressionistic support. For each of the seven beats, please provide a table listing the number of corpus items that engage the topic at all, even partially, and the corresponding item IDs. The promised claim ledger should make this feasible. Without such evidence, readers cannot distinguish 'absent from the corpus' from 'not reported in this review.'
[Section 2 and Section 6] The coding is single-coded and no intercoder reliability is claimed. The central negative claims—'only ten items' and 'none centers a direct interview with a frontier-lab researcher or evaluation engineer'—depend on subjective judgments about 'technical depth' and 'dominant voice.' Please provide a second-coder audit on a random subset (e.g., 20 items), or at least a more precise, pre-registered definition of 'centers' (e.g., the item contains a direct quotation from a named lab researcher/evaluation engineer and that quote is structurally central). The supplement should include the audit trail for these binary claims.

minor comments (4)

[Section 2] The term 'frontier-lab researcher' is used throughout but never defined. Clarify whether this means researchers at OpenAI, Anthropic, Google DeepMind, or a broader set, since the claim 'none centers a direct interview' depends on that boundary.
[Figure 2] The figure mixes two different visualizations (dominant voice and technical depth) in one panel. It would be clearer to split them into two separate figures, and to add item counts directly to the voice bars. The current '0 1 2 3' x-axis for depth is understandable but not fully labeled.
[Section 3.2] The 'Shy Girl' episode and the 'Daggermouth' case are both discussed, but the relationship between them is not made explicit. Consider adding a sentence explaining how the second case illustrates or complicates the chain-of-custody framing introduced earlier.
[References] Several URLs are unattractively long and unformatted, e.g., the ActuaLitté entries and the China News Publishing URL. A consistent citation style with short titles would improve readability. Also, the paper includes references to 2026 events that may not yet be verifiable; please confirm all cited dates and accession dates are accurate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the coverage-gap findings are empirical coding outputs with stated limitations, not reductions to the coding rubric.

full rationale

This is a purposive content analysis, not a derivation that folds its conclusions into its inputs. The central findings—30% risk-framed, 42% mixed, 28% opportunity-framed; ten depth-3 items; no frontier-lab interview at the center of an article—are direct outputs of the author's stated coding of 89 items, with the codebook and supplement disclosed. The coding rubric is not defined in terms of the conclusions, and the conclusions are not used to define the codes. The closest candidate for a circularity concern is the use of the generic depth-3 count ('mechanism, evidence, boundary conditions, and failure modes receive scrutiny') as evidence for the specific 'capability blind spot' (capability elicitation, RAG, prompt injection, agent reliability, inference economics, model drift, provenance, reader research, reproducible workflow evaluation). But the depth construct is broader than the capability-news construct, so the count is at most an imperfect proxy; the paper itself limits the inference: 'technical depth reflects reporting detail rather than whether an article reached the correct policy conclusion.' This is a measurement-validity limitation, not a definitional equivalence. The separate 'none centers a direct interview with a frontier-lab researcher or evaluation engineer' claim rests on dominant-voice coding and is independent of the depth scale. There are no load-bearing self-citations: the references are external (labs, METR, academic studies, C2PA, etc.), and no uniqueness theorem or prior-work ansatz is invoked to force the conclusions. The paper explicitly labels the sample 'an analytic sample, not a census' and disclaims intercoder reliability. Under the stated circularity criteria, no step reduces by construction to its inputs.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The paper introduces no theoretical entities; the central claim rests on methodological choices (sample, coding rubric) that are subjective, disclosed, and auditable only through the promised supplement.

free parameters (2)

technical depth scale = 0-3 ordinal scale
The threshold for depth 3 (mechanism, evidence, boundary conditions, failure modes) is defined by the author; aggregate statistics such as mean depth 1.63 depend on this subjective scale.
corpus inclusion criteria = 89 items
Inclusion depends on the author's judgment that an article is 'primarily concerned with AI's effect on book publishing' (Section 2); no systematic search or duplicate-removal protocol beyond author judgment.

assumptions (3)

domain assumption The purposive sample is representative enough to support field-level generalizations about the trade press.
Section 2 states it is an analytic sample, not a census, and US/UK are overrepresented; the paper's central gap claim extrapolates from this sample.
domain assumption Single-coder coding is sufficiently reliable to support the aggregate claims.
Section 2: 'Coding was performed once by the author; no intercoder reliability is claimed.' The depth and stance distributions are the paper's primary evidence.
domain assumption The technical-depth and stance coding categories are valid constructs for measuring coverage quality.
The 0-3 depth scale assumes that mechanism, evidence, boundary conditions, and failure modes are the right criteria for 'technical scrutiny' (Section 2).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Copyright Is the Headline; Capability Is the Blind Spot: AI Technology in the Book-Publishing Trade Press, November 2025--August 2026." pith.science (2026). https://pith.science/paper/GQEUINZ6

@misc{pith2026260800964,
  author       = {Pith},
  title        = {Pith review of: Copyright Is the Headline; Capability Is the Blind Spot: AI Technology in the Book-Publishing Trade Press, November 2025--August 2026},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GQEUINZ6}},
  note         = {Machine review of arXiv:2608.00964}
}

read the original abstract

This rapid evidence review examines 89 articles about artificial intelligence (AI) and book publishing published from November 1, 2025 through August 1, 2026. The purposive corpus spans English-, Chinese-, German-, French-, Spanish-, Portuguese-, Italian-, and Japanese-language publishing coverage; major-newspaper book coverage; and specialist technology commentators. Each item was coded for topic, stance, technical depth, and dominant voice. The press is neither silent nor simply hostile: 30% of items are risk-framed, 42% mixed, and 28% opportunity-framed. Chinese coverage is markedly operational and opportunity-oriented; specialist commentary is substantially deeper than trade reporting. Yet the corpus still clusters around rights, licensing, governance, reader trust, workflow adoption, and product announcements. Only ten items offer sustained technical scrutiny, and none centers a direct interview with a frontier-lab researcher or evaluation engineer. The central gap is reporting that connects model architecture and evaluation to publishing decisions: capability elicitation, RAG, prompt injection, agent reliability, inference economics, model drift, provenance, reader research, and reproducible workflow evaluation. The report recommends a standing AI beat, recurring lab interviews, a claims ledger, shared test protocols, reader panels, and technical columnists. The supplement supplies the full coded corpus, claim ledger, glossary, and BibTeX database.

Figures

Figures reproduced from arXiv: 2608.00964 by the authors.

**Figure 1.** Primary topic by source stratum. Trade coverage is broad; business/licensing leads, with governance, workflow, and reader experience close behind. Categories were assigned once by the author; counts are descriptive, not population estimates. survey rather than impressions (Josh Bernoff 2026a, 2026b; Josh Bernoff at IBPA PubSpot 2026). Friedman follows governance and generative-engine discoverability; Schmidt foreg… view at source ↗

**Figure 2.** Dominant voice and technical-depth score. Ten articles reach depth 3; the mean is 1.63 on a 0–3 scale. Commentators average 2.14 and trade items 1.54. No sampled article centers a frontier-lab researcher or evaluation engineer. 5 Filling the Gaps 5.1 Recommendations for trade editors 1. Create a standing Future of the Book interview. Monthly, interview one researcher or engineer from Google DeepMind, OpenAI, Anthrop… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

16 extracted references · 14 canonical work pages

[2]

Livres générés par IA : les éditeurs accusent Amazon de parasitisme

https://actualitte.com/article/132344/technologie/les-editeurs-neerlandais- lancent-leur-plateforme-de-licences-ia. . 2026b. “Livres générés par IA : les éditeurs accusent Amazon de parasitisme.” Corpus item C74; primary category: Market structure, April 21, 2026. Accessed August 1, 2026. https://actualitte.com/article/130818/legislation/livres-generes- p...

work page arXiv 2026
[3]

Generative AI floods and dilutes the market for books

https://www.boersenblatt.net/news/boersenverein/wie-verlage-den-einsatz- von-ki-freiwillig-offenlegen-wollen-397617. . 2026a. “Effizienzbooster oder KI-Strategie? Große Unterschiede bei Verlagen.” Corpus item C71; primary category: Production & workflow, February 25, 2026. Accessed August 1, 2026. https://www.boersenblatt.net/news/effizienzbooster- oder-k...

work page Pith review arXiv 2026
[4]

Zheng Weiliang on AI Scientific Publishing Must Shift from Books to Precise Knowledge Delivery

https://chinapublish.cn/xwzx_5852/hydt/202605/t20260525_217672.html. . 2026c. “Zheng Weiliang on AI Scientific Publishing Must Shift from Books to Precise Knowledge Delivery.” Corpus item C61; primary category: Technology, February 12, 2026. Accessed August 1, 2026. https://book.cctv.cn/2026/02/12/ ARTIsGLSfE61lW8P3EzQVrfm260212.shtml. China Publishing & ...

work page 2026
[5]

IntheAIEraDoWeStillNeedDeepReading

https://www.cbbr.com.cn/contents/533/106448.html. .2026e.“IntheAIEraDoWeStillNeedDeepReading.”CorpusitemC67;primary category: Reader experience, June 18, 2026. Accessed August 1, 2026. https://www. cbbr.com.cn/contents/533/110151.html. .2026f.“WorldChildren’sBookForumDiscussesPublishingStrategyintheAIEra.” Corpus item C68; primary category: Reader experie...

arXiv doi:10.48550/arxiv.2604.08304 2026
[6]

While Writers Worry About AI Many Have Em- bracedIt

Accessed August 1, 2026. https://bernoff.com/blog/ingram-lets-publishers- opt-out-of-sales-to-ai-companies-to-bad-that-wont-work. Josh Bernoff at IBPA PubSpot. 2026. “While Writers Worry About AI Many Have Em- bracedIt.”CorpusitemC64;primarycategory:Governance,June5,2026.Accessed August 1, 2026. https://pubspot.ibpa-online.org/article/while-writers-worry-...

arXiv 2026
[8]

Chicken Soup for the Soul Sues AI Firms for Copyright Infringement

https://www.publishersweekly.com/pw/by-topic/digital/content-and- e-books/article/99705-the-stealthy-startup-promising-better-ai-publishing- licensing-deals.html. . 2026e. “Chicken Soup for the Soul Sues AI Firms for Copyright Infringement.” Corpus item C25; primary category: Legal & regulatory, March 20, 2026. Accessed August 1, 2026. https://www.publish...

work page 2026
[9]

Interactive AI Features in E-books Audiobooks Drive Debate

https://www.publishersweekly.com/pw/by-topic/industry-news/publisher- news/article/99453-ingram-offers-to-not-sell-books-to-ai-companies.html. . 2026j. “Interactive AI Features in E-books Audiobooks Drive Debate.” Corpus itemC08;primarycategory:Readerexperience,January5,2026.AccessedAugust1,

work page 2026
[10]

Publishers Authors File Class Action Lawsuit Against Google

https://www.publishersweekly.com/pw/by-topic/international/international -book-news/article/99381-amazon-elevenlabs-new-digital-interaction-features- drive-debate.html. . 2026k. “Publishers Authors File Class Action Lawsuit Against Google.” Corpus item C47; primary category: Legal & regulatory, July 13, 2026. Accessed August 1,

work page 2026

Show all 16 references

[11]

Publishers File Lawsuit Against Meta Mark Zuckerberg

https://www.publishersweekly.com/pw/by-topic/digital/copyright/article/ 100820-new-lawsuit-aims-to-stop-google-from-copyright-infringement-in- creating-ai-models.html. . 2026l. “Publishers File Lawsuit Against Meta Mark Zuckerberg.” Corpus item C34; primary category: Legal & r...

2026
[12]

The Future of the Italian Market: Vouchers Booksellers and Capitalizing onAI

https://publishingperspectives.com/2026/03/exploring-new-revenue- opportunities-through-licensing/. . 2026b. “The Future of the Italian Market: Vouchers Booksellers and Capitalizing onAI.”CorpusitemC14;primarycategory:Business&licensing,February4,2026. Accessed August 1, 2026....

2026
[14]

Soon publishers won’t stand a chance: struggle to detect AI-written books

Accessed August 1, 2026. https://www.theguardian.com/books/2026/mar/ 13/grammarly-removes-ai-expert-review-feature-mimicking-writers-after- backlash. . 2026b. “Soon publishers won’t stand a chance: struggle to detect AI-written books.”CorpusitemC30;primarycategory:Provenance,M...

2026
[15]

A.I. Is Writing Fiction. Publishers Are Unprepared

https://www.japantimes.co.jp/culture/2026/07/03/books/haruki-murakami- new-book/. The New York Times. 2026a. “A.I. Is Writing Fiction. Publishers Are Unprepared.” Corpus itemC24;primarycategory:Provenance,March19,2026.AccessedAugust1,2026. https://www.nytimes.com/2026/03/19/bo...

2026
[16]

Who wrote this? Evaluating the reliability of AI detection tools in higher education

https://www.nytimes.com/2026/03/25/opinion/shy-girl-ai-publishing.html. UK Department for Science, Innovation and Technology and Intellectual Property Of- fice. 2026.Report and Impact Assessment on Copyright and Artificial Intelligence. March 18, 2026. Accessed August 1, 2026....

2026
[1315]

Licensing for AI What Should Book Publish- ers and Authors Do

https://doi.org/10.1016/j.compedu.2026.105616. http://dx.doi.org/10.1016/j. compedu.2026.105616. Svenska Foerlaeggarefoereningen. 2026.Svenskarna vill inte laesa AI-genererade boecker. Kantar Media survey, n=1,031. April 23, 2026. Accessed August 1, 2026. https: //forlaggare.s...

2026
[2025]

Introducing GPT-5.2

Accessed August 1, 2026. https://openai.com/index/evals-drive-next- chapter-of-ai/. . 2025b. “Introducing GPT-5.2.” Vendor release; capability claims are self-reported, December 11, 2025. Accessed August 1, 2026. https://openai.com/index/introduci ng-gpt-5-2/. . 2026a. “A Shar...

2026
[2026]

LeséditeursnéerlandaislancentleurplateformedelicencesIA

Accessed August 1, 2026. https://www.actualidadeditorial.com/datasets-el- nuevo-valor-economico-del-fondo-editorial/. ActuaLitté.2026a.“LeséditeursnéerlandaislancentleurplateformedelicencesIA.”Corpus itemC75;primarycategory:Business&licensing,June29,2026.AccessedAugust1,

2026

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [2]

Livres générés par IA : les éditeurs accusent Amazon de parasitisme

https://actualitte.com/article/132344/technologie/les-editeurs-neerlandais- lancent-leur-plateforme-de-licences-ia. . 2026b. “Livres générés par IA : les éditeurs accusent Amazon de parasitisme.” Corpus item C74; primary category: Market structure, April 21, 2026. Accessed August 1, 2026. https://actualitte.com/article/130818/legislation/livres-generes- p...

work page arXiv 2026

[2] [3]

Generative AI floods and dilutes the market for books

https://www.boersenblatt.net/news/boersenverein/wie-verlage-den-einsatz- von-ki-freiwillig-offenlegen-wollen-397617. . 2026a. “Effizienzbooster oder KI-Strategie? Große Unterschiede bei Verlagen.” Corpus item C71; primary category: Production & workflow, February 25, 2026. Accessed August 1, 2026. https://www.boersenblatt.net/news/effizienzbooster- oder-k...

work page Pith review arXiv 2026

[3] [4]

Zheng Weiliang on AI Scientific Publishing Must Shift from Books to Precise Knowledge Delivery

https://chinapublish.cn/xwzx_5852/hydt/202605/t20260525_217672.html. . 2026c. “Zheng Weiliang on AI Scientific Publishing Must Shift from Books to Precise Knowledge Delivery.” Corpus item C61; primary category: Technology, February 12, 2026. Accessed August 1, 2026. https://book.cctv.cn/2026/02/12/ ARTIsGLSfE61lW8P3EzQVrfm260212.shtml. China Publishing & ...

work page 2026

[4] [5]

IntheAIEraDoWeStillNeedDeepReading

https://www.cbbr.com.cn/contents/533/106448.html. .2026e.“IntheAIEraDoWeStillNeedDeepReading.”CorpusitemC67;primary category: Reader experience, June 18, 2026. Accessed August 1, 2026. https://www. cbbr.com.cn/contents/533/110151.html. .2026f.“WorldChildren’sBookForumDiscussesPublishingStrategyintheAIEra.” Corpus item C68; primary category: Reader experie...

arXiv doi:10.48550/arxiv.2604.08304 2026

[5] [6]

While Writers Worry About AI Many Have Em- bracedIt

Accessed August 1, 2026. https://bernoff.com/blog/ingram-lets-publishers- opt-out-of-sales-to-ai-companies-to-bad-that-wont-work. Josh Bernoff at IBPA PubSpot. 2026. “While Writers Worry About AI Many Have Em- bracedIt.”CorpusitemC64;primarycategory:Governance,June5,2026.Accessed August 1, 2026. https://pubspot.ibpa-online.org/article/while-writers-worry-...

arXiv 2026

[6] [8]

Chicken Soup for the Soul Sues AI Firms for Copyright Infringement

https://www.publishersweekly.com/pw/by-topic/digital/content-and- e-books/article/99705-the-stealthy-startup-promising-better-ai-publishing- licensing-deals.html. . 2026e. “Chicken Soup for the Soul Sues AI Firms for Copyright Infringement.” Corpus item C25; primary category: Legal & regulatory, March 20, 2026. Accessed August 1, 2026. https://www.publish...

work page 2026

[7] [9]

Interactive AI Features in E-books Audiobooks Drive Debate

https://www.publishersweekly.com/pw/by-topic/industry-news/publisher- news/article/99453-ingram-offers-to-not-sell-books-to-ai-companies.html. . 2026j. “Interactive AI Features in E-books Audiobooks Drive Debate.” Corpus itemC08;primarycategory:Readerexperience,January5,2026.AccessedAugust1,

work page 2026

[8] [10]

Publishers Authors File Class Action Lawsuit Against Google

https://www.publishersweekly.com/pw/by-topic/international/international -book-news/article/99381-amazon-elevenlabs-new-digital-interaction-features- drive-debate.html. . 2026k. “Publishers Authors File Class Action Lawsuit Against Google.” Corpus item C47; primary category: Legal & regulatory, July 13, 2026. Accessed August 1,

work page 2026