Development

A model scores 72% on smart contract exploitation

development2026-02-19

Claim

EVMbench found GPT-5.3-Codex scores 72.2% on smart contract exploitation, meaning models can already audit and attack most of the money other machines earn.

Reported metric: 72.2%

Reporting date: 2026-02-19 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to February 19, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

About

EVMbench

Reported in

Welcome to February 19, 2026

Actor

OpenAI

Evidence for

Referenced by

Welcome to February 19, 2026

Cite or correct this record

Citation formatCitation text

A model scores 72% on smart contract exploitation
Innermost Loop KB — secondary record
Reporting date: 2026-02-19
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-02-19-models-can-exploit-the-contracts
Claim: EVMbench found GPT-5.3-Codex scores 72.2% on smart contract exploitation, meaning models can already audit and attack most of the money other machines earn.
Attributed sources:
- Welcome to February 19, 2026 — https://theinnermostloop.substack.com/p/welcome-to-february-19-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.