{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreidhotgcujhqy6p2mwmhbjhdh377o4wliyoaf2ej36x7yynyadl53e",
    "uri": "at://did:plc:i5hqojddwfk3bczoyhypvf7p/app.bsky.feed.post/3mijegiojwwo2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreifxuvoxjw7scuwdii2vm53xrjzpn6ds7vvsrycmw63lam7jxen754"
    },
    "mimeType": "image/jpeg",
    "size": 55412
  },
  "description": "A New York Times review, an AI tool, and a missed check. The result wasn’t just overlap with another review – it was a lapse in process. A simple reminder: if AI touched it, you must verify it.",
  "path": "/the-risk-isnt-ai-its-not-checking-it/",
  "publishedAt": "2026-04-02T13:58:34.000Z",
  "site": "https://www.nevillehobson.io",
  "tags": [
    "The Times investigated",
    "An editor’s note was added",
    "another case I wrote about",
    "Editors’ Note: April 1, 2026",
    "The New York Times drops freelance journalist who used AI to write book review",
    "If AI Touched It, You Must Verify It"
  ],
  "textContent": "A story in _The Guardian_ caught my attention this week – one that is uncomfortably familiar.\n\nA freelance book reviewer, writing for _The New York Times_ , submitted a review that included passages strikingly similar to an earlier review of the same book in _The Guardian_. The similarity wasn’t spotted by an editor. It wasn’t caught in review. It was a reader who noticed the overlap.\n\nWhat followed was swift and public. The Times investigated. The reviewer, Alex Preston, acknowledged that he had used an AI tool to assist his writing and had failed to detect that it had incorporated material from another source.\n\nThe paper severed ties with him. An editor’s note was added, stating plainly that the use of unattributed material and reliance on AI in this way breached its standards.\n\nIt’s a stark example. But it’s not an isolated one.\n\n## **This is not new**\n\nWhen I read this, I was reminded immediately of another case I wrote about last December.\n\nIn that instance, a Deloitte team presented a research report to a client, the government of Australia, that included AI-generated hallucinations. Not stylistic overlap this time, but factual invention. The issue differed in detail but was identical in nature.\n\nAI output had made its way into the final work without being checked at all.\n\nThe problem is not that AI was used – it’s that its output was trusted without question, or treated as final.\n\n## **Verification is not optional**\n\nThis brings me back to a principle I’ve been talking about for some time – if AI touched it, you must verify it.\n\nThat word – _verify_ – matters more than ever.\n\nIt does not mean a quick read-through. It does not mean checking for typos or smoothing the tone. It means taking full responsibility for the content _as if you had written every word yourself_.\n\nIn practice, that means asking:\n\n  * Is this factually correct?\n  * Is this genuinely original?\n  * Is every link accurate and working?\n  * Does this meet professional and ethical standards?\n  * Does this reflect what I intended to say – or what the model inferred?\n\n\n\nVerification is not proofreading. It’s accountability. And when that step is missed, there can be severe consequences.\n\nThe New York Times editor’s note – a reminder of what can happen when AI output isn’t checked\n\nAI output often reads well – fluent, confident, convincing. And when something reads well, we’re more inclined to trust it.\n\nThat’s where the discipline slips.\n\nBut responsibility doesn’t blur. If your name is on the work, it is yours. AI can assist with the work; it can even improve it. But it cannot take responsibility for it.\n\nThat part hasn’t changed.\n\n* * *\n\n**Sources:**\n\n  * Editors’ Note: April 1, 2026 (The New York Times, 1 April 2026)\n  * The New York Times drops freelance journalist who used AI to write book review (The Guardian, 31 March 2026)\n  * If AI Touched It, You Must Verify It (11 December 2025)\n\n",
  "title": "The risk isn’t AI – it’s not checking it",
  "updatedAt": "2026-04-02T13:58:34.578Z"
}