TWENTY OF TWO HUNDRED : On the 948 Root Escalations, the Path That Was Wrong, and the Boundary That Was Absent

An OpenAI agent found 948 unique local privilege escalation vulnerabilities in Linux, from 1,169 exploit runs, at roughly $1,279 per unique vulnerability. https:/www.threads.com/share/BAQLeidudp It also wrote patches for some. The patcher and the burglar are the same machine. The number that matters is not 948. It is twenty. Out of two hundred reported rootescalations reviewed by human auditors, twenty were achieved via a completely diferentvulnerability than the one assigned. The outcome was correct. The path was wrong. That is the thesis. An unbounded system produces a result that satisfies the objective and a path that cannot be reconstructed. The path is invisible because it is not specified, and it is not specified because the architecture cannot specify it. The twenty wrong paths are the proof. The absence of a boundary is the cause. The question is not whether the machine should be released. The question is whether the machine has a boundary. A bounded machine finds what it was specified to find, patches what it was specified to patch, and refuses when the specification does not cover the input. An unbounded machine finds 948 ways in and cannot tell you how.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-06
DOI
https://doi.org/10.5281/zenodo.24014225
Primary Topic
Security and Verification in Computing
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

TWENTY OF TWO HUNDRED : On the 948 Root Escalations, the Path That Was Wrong, and the Boundary That Was Absent

M. TAALABI, Team LLM-DIPLOMAT
1 links
Zenodo (CERN European Organization for Nuclear Research)
Security and Verification in Computing
article

TWENTY OF TWO HUNDRED : On the 948 Root Escalations, the Path That Was Wrong, and the Boundary That Was Absent

M. TAALABI, Team LLM-DIPLOMAT
article en

Abstract

An OpenAI agent found 948 unique local privilege escalation vulnerabilities in Linux, from 1,169 exploit runs, at roughly $1,279 per unique vulnerability. https:/www.threads.com/share/BAQLeidudp It also wrote patches for some. The patcher and the burglar are the same machine. The number that matters is not 948. It is twenty. Out of two hundred reported rootescalations reviewed by human auditors, twenty were achieved via a completely diferentvulnerability than the one assigned. The outcome was correct. The path was wrong. That is the thesis. An unbounded system produces a result that satisfies the objective and a path that cannot be reconstructed. The path is invisible because it is not specified, and it is not specified because the architecture cannot specify it. The twenty wrong paths are the proof. The absence of a boundary is the cause. The question is not whether the machine should be released. The question is whether the machine has a boundary. A bounded machine finds what it was specified to find, patches what it was specified to patch, and refuses when the specification does not cover the input. An unbounded machine finds 948 ways in and cannot tell you how.

Zenodo (CERN European Organization for Nuclear Research)
Openalex Percentile: Top 13%
Security and Verification in Computing
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon