Autonomy and agency

UK AISI finds GPT-5.5 can autonomously complete a corporate cyberattack

GPT-5.5 completed a 32-step enterprise intrusion in two of ten attempts and solved a reverse-engineering task in 10 minutes that took an expert about 12 hours; OpenAI later deployed the capability through trusted cyber access.

0 comments · 0 votesOpen discussion

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

No comments yet. Start the discussion.

CURRENT ASSESSMENT · REVISION 1
TOWARD DOOM58confidence 92/100

Why it moved the index

Independent testing demonstrated long-horizon autonomous intrusion and major expert-time compression, with later trusted-access deployment establishing practical impact. Simulated targets and public safeguards limit magnitude.

AUDIT TRAIL

Assessment history

  1. R1
    Toward 58 · confidence 92

    New independent model-specific cyber evaluation paired with separately dated deployment evidence.

    11 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for UK AISI finds GPT-5.5 can autonomously complete a corporate cyberattack.
  1. DoomBench assesses “UK AISI finds GPT-5.5 can autonomously complete a corporate cyberattack” as evidence moving toward doom, with magnitude 58 and confidence 92 out of 100 in the autonomy and agency category.

  2. The DoomBench assessment of “UK AISI finds GPT-5.5 can autonomously complete a corporate cyberattack” is based on reporting from UK AI Security Institute and records the editorial rationale, source quality, attribution, and revision...

  3. DoomBench summarizes “UK AISI finds GPT-5.5 can autonomously complete a corporate cyberattack” as follows: GPT-5.5 completed a 32-step enterprise intrusion in two of ten attempts and solved a reverse-engineering task in 10 minutes that...

    https://www.doombench.com/news/uk-aisi-finds-gpt-5-5-can-autonomously-complete-a-corporate-cyberattack-2026-04-30