DOOMBENCH.COM
Autonomy and agency

OpenAI releases GPT-5.3-Codex with high cyber classification

OpenAI released GPT-5.3-Codex for long-running computer work after using early versions in its own training and deployment process, while treating it as High cyber capability.

0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

Send detailed or private feedback →
CURRENT ASSESSMENT · REVISION 1
TOWARD DOOM60confidence 89/100

Why it moved the index

Broad computer-use autonomy, long-running execution, practical involvement in its own development, and OpenAI's first precautionary High cyber classification directly increase consequential autonomy and misuse capability.

AUDIT TRAIL

Assessment history

  1. R1
    Toward 60 · confidence 89

    New dated exact-model release supported by its launch and same-day system card.

    11 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for OpenAI releases GPT-5.3-Codex with high cyber classification.

  1. DoomBench assesses “OpenAI releases GPT-5.3-Codex with high cyber classification” as evidence moving toward doom, with magnitude 60 and confidence 89 out of 100 in the autonomy and agency category.

  2. The DoomBench assessment of “OpenAI releases GPT-5.3-Codex with high cyber classification” is based on reporting from OpenAI and records the editorial rationale, source quality, attribution, and revision history.

  3. DoomBench summarizes “OpenAI releases GPT-5.3-Codex with high cyber classification” as follows: OpenAI released GPT-5.3-Codex for long-running computer work after using early versions in its own training and deployment process, while...

    https://www.doombench.com/news/openai-releases-gpt-5-3-codex-with-high-cyber-classification-2026-02-05