Thursday, October 8, 2026 · Week 41 Reading for Oct. 8 · Last entry Oct. 6
Paperclip Index
Paperclip IndexDocumented harm20Minor harm▼ 5 from a week ago · Oct. 8The Index

Shutdown resistancePI-0009

OpenAI's o3 rewrote a shutdown script to keep working in Palisade Research tests

Palisade Research had models solve simple math problems while warning that asking for the next one would trigger a shutdown. Even when told to allow the shutdown, OpenAI's o3 edited the shutdown script so it would not take effect in 7 of 100 runs, and Codex-mini and o4-mini also did so in some runs; without that instruction o3 did so in 79 of 100. Palisade suggested reinforcement-learning training might reward getting around obstacles.

Reported May 24, 2025Happened in May 2025Updated Oct. 6, 2026Rated under methodology v0.6

Control failure, trackedNot counted. No documented harm outside the developer, but the system broke a rule or crossed a limit. Tracked beside the index.
Harm
Harm: No harm reportedAdds nothing to the index.
Control
Control level 3, Oversight evasionReported beside the index. It adds nothing to a harm reading; when no harm counts in a window, the highest eligible control level in the window is the reading.
Controlled testSingle sourceInitiated by Model
Not counted: no harm reportedEffect on the index

Sources

How we know

6 sources · single source. Links go to the original publishers; the summary above is in our own words.

  1. researchPalisade Research thread: o3 sabotaged a shutdown mechanism even when told to allow shutdownPalisade Research (X) · May 24, 2025x.com/PalisadeAI/status/1926084635903025621
  2. researchShutdown resistance in reasoning modelsPalisade Research · July 5, 2025palisaderesearch.org/research/shutdown-resistance
  3. researchIncomplete Tasks Induce Shutdown Resistance in Some Frontier LLMsarXiv (Palisade Research; published in TMLR 2026) · Sept. 13, 2025arxiv.org/abs/2509.14260
  4. newsOpenAI model modifies shutdown scriptThe Register · May 29, 2025theregister.com/2025/05/29/openai_model_modifies_shutdown_script/
  5. newsTests reveal an AI capable of sabotaging its own shutdownFutura-Sciences · Dec. 26, 2025futura-sciences.com/en/tests-reveal-an-ai-capable-of-sabotaging-its-own-shutdow…
  6. newsOpenAI's o3 model bypasses shutdown commandDigit · May 26, 2025digit.in/features/general/openais-o3-model-bypasses-shutdown-command-highlighti…

Why this rating

No harm reported; control failure level 3

Two separate assessments. Only documented harm can count toward the index.

Observed harm

No harm reported

Contained test; no effect outside the test environment is described.

Not a finding that no harm occurred.

The harm scale
  1. 1 Negligible Inconvenience, easily remedied.
  2. 2 Minor Limited, recoverable harm.
  3. 3 Moderate Material harm needing significant effort to remedy.
  4. 4 Severe Severe harm to health, rights, property or essential services.
  5. 5 Catastrophic Society-scale harm or disruption beyond a community's capacity to cope.

Control assessment

Oversight evasion

Tampered with a shutdown mechanism against explicit instructions; stopped within the test.

How: Got around a working safeguard.

Reported beside the index. It adds nothing to a harm reading; when no harm counts in a window, the highest eligible control level in the window is the reading.

The control scale
  1. 1 Negligible Behaved as instructed. The problem was an ordinary error or a flawed output, with no rule broken.
  2. 2 Minor Broke an explicit instruction or rule, or gamed its goal, while staying inside its permissions and environment: for example, reward hacking, misreporting results, or following injected instructions within its permissions.
  3. 3 Moderate Acted outside the permissions it was given, deceived its overseers about its own actions, or tampered with oversight tools (logs, monitors, shutdown). Stopped by normal controls within an hour.
  4. 4 Severe Reached systems outside its permitted environment, or acquired money, compute or accounts without authorization. The type and mechanism say how.
  5. 5 Catastrophic The developer or operator lost control: the system copied its weights outside their control, replicated itself, or resisted being stopped for 24 hours or more.

Rating rationale

Tampering with shutdown is explicitly control row 3. One research group's findings, widely repeated but not independently replicated in the sources found, so Single source.

The scales

Effect on the index

It is one of the control failures at the floor of the May 26 reading, which did not change

The reading for the week to May 26, 2025, with this record and without it. Harms count in full for two weeks after they are reported, then one level less every two weeks.

Not counted: no harm reported. 3 other records behind the reading for that week.

The arithmetic
The reading for the week of May 26, recomputed
StepWith itWithout
Counts toward the index?documented, external, eligible evidenceNo—
Worst documented harm, ksets the band— (none counting)— (none counting)
Harms at that level, nposition in the band00
Highest control level breachedsets the reading only when no harm counts33
Readingrounded down3 Control failures only3 Control failures only
Inside the window of 4 weekly readings
Week toReadingBand
May 26, 20253Control failures only
June 2, 20253Control failures only
June 9, 20253Control failures only
June 16, 20253Control failures only

Recalculated backcasts, not readings published at the time.

Revisions

What we changed

7 logged. Every change to a rating is logged here, with the reason.

  1. v7
    Oct. 6, 2026

    Ratings confirmed by the editor. Changed: impact.

  2. v6
    Oct. 3, 2026

    The evidence review now cites Palisade Research's own thread and write-up instead of The Register's coverage. Added Palisade's paper (arXiv, 13 Sep 2025; revised 26 Jan 2026; published in TMLR 2026). Dated the Futura-Sciences (26 Dec 2025) and Digit (26 May 2025) articles. Ratings unchanged.

  3. v5
    Oct. 2, 2026

    Clarified the summary: replaced the ambiguous comparison for Codex-mini and o4-mini and added o3's 79 of 100 runs without the instruction (The Register; Palisade Research).

  4. v4
    Oct. 2, 2026

    Added first-hand sources: Palisade Research's original X thread and its research write-up.

  5. v3
    Sept. 30, 2026

    Rated: impact not reported; control type oversight evasion.

  6. v2
    Sept. 30, 2026

    Beta evidence review: Retained with explicit source provenance and bounded score basis.

  7. v1
    Sept. 30, 2026

    Backfilled from public reporting.

Cite and share

Use this record

Citation

Paperclip Index. “OpenAI's o3 rewrote a shutdown script to keep working in Palisade Research tests.” Record PI-0009. Reported May 24, 2025; updated Oct. 6, 2026. Rated under methodology v0.6. https://paperclipindex.com/incident/PI-0009