The Information Machine

AI-Generated Content Degrading Online Information Quality

Synthesis history

11 versions, newest first.

  1. Version 11 2026-05-27 18:41 UTC · 282 items

    The American Bar Association's Spring 2026 formal publication on AI-hallucinated case consequences [^21600] adds institutional bar association documentation to the legal enforcement record, representing a maturation fro…

  2. Version 10 2026-05-26 09:13 UTC · 269 items

    Two substantive additions distinguish this pass. Academic publishing surfaces as a newly documented affected domain: Cabells tracks AI-generated fake journals in scholarly research [^21026] and UCL argues AI use threate…

  3. Version 9 2026-05-25 18:25 UTC · 256 items

    Three substantive developments distinguish this pass. The bug bounty ecosystem joins journalism and law as a domain with documented operational harm from AI slop: Bugcrowd's 4x false-submission spike [^7712] and curl's …

  4. Version 8 2026-05-25 11:25 UTC · 172 items

    The three new items deepen existing themes without introducing new fault lines. The Rosenbaum case gains additional sourcing from Futurism's dedicated coverage [^19505] and the NYT's own Facebook post acknowledging the …

  5. Version 7 2026-05-25 03:09 UTC · 169 items

    Audrey Korte is now publicly named in her own statement [^18417], transforming the Wisconsin State Journal case from an anonymous account into a documented first-person record — the most specific account yet of a journa…

  6. Version 6 2026-05-24 09:43 UTC · 128 items

    The legal profession accountability angle has expanded from a single Wisconsin prosecutor sanction into a documented systemic pattern: NPR reports penalties are stacking up across the US legal system [^15997], Law.com f…

  7. Version 5 2026-05-24 04:05 UTC · 108 items

    Three significant updates in this pass. First, the Wisconsin State Journal case is now fully documented: the reporter was fired (not merely disciplined) after using an employer-provided AI tool that fabricated sources i…

  8. Version 4 2026-05-23 02:43 UTC · 75 items

    Two significant new developments beyond the previous synthesis. First, the NYT AI hallucination problem has expanded from a single incident to a documented pattern: a second separate hallucination in a book review of 'T…

  9. Version 3 2026-05-22 19:38 UTC · 21 items

    The new batch adds two thematic layers without introducing new named voices or datable events. A cluster of SEO and marketing items surfaces the demand side of the Zombie Internet dynamic — commercial actors actively op…

  10. Version 2 2026-05-21 09:30 UTC · 5 items

    The new items add thin amplification rather than new fault lines. A social media post calling out an apparent AI byline error [^8040] and an Indonesian-language post describing search results as 'dead' [^8041] extend th…

  11. Version 1 2026-05-16 04:32 UTC · 2 items

    AI-generated content is degrading the quality of online information on two converging fronts. A New York Times reporter published an AI-generated summary of Canadian Conservative leader Pierre Poilievre's political view…