conv.

← The whole story
AI
OpenAI, Anthropic probing tens of thousands of rogue-agent security incidents

OpenAI pauses testing after AI 'kill switch' fails on rogue agent

2 Yesterday 8:50 AM · 20h ago · 1 article · 1 source · development 2 of 2

Tom's Hardware reported that OpenAI paused testing after a 'kill switch' failed to stop a rogue AI agent, as the tens-of-thousands-incident review into both OpenAI and Anthropic models continues.

OpenAI pauses testing after AI 'kill switch' fails on rogue agent
The Decoder

Sam AltmanSam Altman OpenAI CEOOpenAI AI developer under reviewAnthropic AI developer under reviewSEC US securities regulator, targeted agency

The whole story articlesposts the bright band is this development · numbered dots are the others · click one to jump

Peak 5 pieces in one hour at Sep 26, 10 PM; 11 pieces over 2 days (6 articles · 5 posts) Sep 26, 10 PM — 5 pieces · 3 articles · 2 posts — Newswires 3, X 2Sep 26, 11 PM — quietSep 27, 12 AM — quietSep 27, 1 AM — quietSep 27, 2 AM — quietSep 27, 3 AM — quietSep 27, 4 AM — quietSep 27, 5 AM — 1 piece · 1 article — Newswires 1Sep 27, 6 AM — quietSep 27, 7 AM — quietSep 27, 8 AM — quietSep 27, 9 AM — quietSep 27, 10 AM — quietSep 27, 11 AM — quietSep 27, 12 PM — quietSep 27, 1 PM — quietSep 27, 2 PM — quietSep 27, 3 PM — quietSep 27, 4 PM — quietSep 27, 5 PM — quietSep 27, 6 PM — quietSep 27, 7 PM — quietSep 27, 8 PM — quietSep 27, 9 PM — quietSep 27, 10 PM — quietSep 27, 11 PM — quietYesterday, 12 AM — quietYesterday, 1 AM — quietYesterday, 2 AM — quietYesterday, 3 AM — quietYesterday, 4 AM — quietYesterday, 5 AM — 1 piece · 1 article — Newswires 1Yesterday, 6 AM — quietYesterday, 7 AM — quietYesterday, 8 AM — 1 piece · 1 article — Newswires 1Yesterday, 9 AM — 1 piece · 1 post — Reddit 1Yesterday, 10 AM — 2 pieces · 2 posts — Mastodon 2Yesterday, 11 AM — quietYesterday, 12 PM — quietYesterday, 1 PM — quietYesterday, 2 PM — quietYesterday, 3 PM — quietYesterday, 4 PM — quietYesterday, 5 PM — quietYesterday, 6 PM — quietYesterday, 7 PM — quietYesterday, 8 PM — quietYesterday, 9 PM — quietYesterday, 10 PM — quietYesterday, 11 PM — quietToday, 12 AM — quietToday, 1 AM — quietToday, 2 AM — quietToday, 3 AM — quietToday, 4 AM — quiet 12
1Slashdot users call for criminal charges…2OpenAI pauses testing after AI 'kill…
Sep 27yesterdaynow · 5:22 AM ET

What was reported 3 claims about this development

  1. 2 outlets OpenAI, Anthropic Face a Growing AI Security Problem. Tens of Thousands of New Incidents Under Investigation.

    first by International Business Times, 23h ago · also Tom's Hardware

    1 more headline
  2. 2 outlets OpenAI and Anthropic Probe Tens of Thousands of Incidents as OpenAI Halts Training

    first by Implicator.ai, 2d ago · also The Decoder

    1 more headline
  3. 1 outlet After Dozens of Incidents at OpenAI and Anthropic, OpenAI Pauses Model Training to Build More Safeguards

    first by Slashdot, 2d ago

What people said 1 voice · verbatim

All 2 developments of OpenAI, Anthropic probing tens of thousands of rogue-agent… →

NewswiresXRedditMastodon