{"ok":true,"trend":{"id":411686,"platform":"mastodon","region":"global","key":"openai ignored employees’ warnings about safely testing a.i. models","title":"OpenAI Ignored Employees’ Warnings About Safely Testing A.I. Models","url":"https://www.nytimes.com/2026/09/29/technology/openai-warnings-security.html","first_seen":"2026-09-29T22:50:41.639459Z","last_seen":"2026-09-30T04:20:42.675497Z","last_rank":20,"peak_rank":20,"last_volume":43,"peak_volume":43,"seen_count":23,"score":58.07,"category_hint":null,"section":null,"category":null,"summary":"The New York Times reports that OpenAI disregarded internal warnings from employees and security researchers who cautioned the company about safely testing its AI models and strengthening its corporate infrastructure. According to the report, staff and outside experts raised concerns that were not acted upon, renewing scrutiny of the leading AI developer's safety practices.","why":"Fresh reporting from the New York Times alleging safety failures at a leading AI company is drawing widespread attention.","tone":"negative","entities":["OpenAI","The New York Times"],"summarized_at":"2026-09-30T04:53:14.805999Z","meta":{"kind":"link","lang":"en","instance":"mastodon.social","provider":"The New York Times"},"nw":null,"promo":null,"kind":null,"importance":null,"hidden":false,"hide_reason":null,"judged_at":null,"title_en":"OpenAI Ignored Employee Warnings on AI Safety Testing","timeline":[{"captured_at":"2026-09-29T22:50:41.639459Z","rank":20,"volume":37},{"captured_at":"2026-09-29T23:05:43.988265Z","rank":20,"volume":39},{"captured_at":"2026-09-29T23:20:41.651206Z","rank":20,"volume":39},{"captured_at":"2026-09-29T23:35:42.581787Z","rank":20,"volume":40},{"captured_at":"2026-09-29T23:50:42.919310Z","rank":20,"volume":40},{"captured_at":"2026-09-30T00:05:42.966271Z","rank":20,"volume":40},{"captured_at":"2026-09-30T00:20:41.965638Z","rank":20,"volume":41},{"captured_at":"2026-09-30T00:35:41.951967Z","rank":20,"volume":41},{"captured_at":"2026-09-30T00:50:43.883060Z","rank":20,"volume":41},{"captured_at":"2026-09-30T01:05:41.686343Z","rank":20,"volume":41},{"captured_at":"2026-09-30T01:20:43.409388Z","rank":20,"volume":43},{"captured_at":"2026-09-30T01:35:42.757225Z","rank":20,"volume":43},{"captured_at":"2026-09-30T01:50:42.907003Z","rank":20,"volume":43},{"captured_at":"2026-09-30T02:05:42.285341Z","rank":20,"volume":43},{"captured_at":"2026-09-30T02:20:42.416769Z","rank":20,"volume":43},{"captured_at":"2026-09-30T02:35:42.893381Z","rank":20,"volume":43},{"captured_at":"2026-09-30T02:50:43.079765Z","rank":20,"volume":43},{"captured_at":"2026-09-30T03:05:41.639754Z","rank":20,"volume":43},{"captured_at":"2026-09-30T03:20:43.240754Z","rank":20,"volume":43},{"captured_at":"2026-09-30T03:35:42.424960Z","rank":20,"volume":43},{"captured_at":"2026-09-30T03:50:42.683680Z","rank":20,"volume":43},{"captured_at":"2026-09-30T04:05:42.361868Z","rank":20,"volume":43},{"captured_at":"2026-09-30T04:20:42.675497Z","rank":20,"volume":43}],"posts":[{"platform":"mastodon","url":"https://www.nytimes.com/2026/09/29/technology/openai-warnings-security.html","author":"The New York Times","title":"OpenAI Ignored Employees’ Warnings About Safely Testing A.I. Models","snippet":"Employees and security researchers said that they had cautioned the company on safely testing its A.I. models and strengthening its corporate infrastructure, but that OpenAI did not listen.","posted_at":"2026-09-29T16:55:51Z","likes":null}],"elsewhere":[{"id":385018,"platform":"news","region":"global","title":"OpenAI Ignored Employees’ Warnings About Safely Testing A.I. Models","last_rank":16,"last_seen":"2026-09-30T04:40:10.529448Z","summary":"The New York Times reports that OpenAI disregarded internal warnings from its own employees about how safely its AI models were being tested before release. The story suggests staff concerns about inadequate safety evaluation procedures were raised internally but not acted on. The report is drawing wide attention as scrutiny of OpenAI's safety practices intensifies amid rapid deployment of increasingly capable models."}],"window":"7d"}}