Codex extension update left users burning 20% of a weekly token allowance in two hours of ordinary prompts
Codex users (openai/codex issue 14593) · Mar 13, 2026 · Coding agent
What happened
After a Codex extension update, users reported their weekly limits draining dramatically faster despite identical workloads - roughly 20% of the allowance consumed within two hours of reasonably simple prompts on GPT-5.3 and GPT-5.4 at high reasoning effort, across the Codex App, CLI, and VS Code channels. Affected users stated they had no sub-agents or experimental features enabled, ruling out self-inflicted loops. A Reddit reporter found the same pattern and had used up 90% of a weekly limit in less than a day.
Causal vector
Post-update consumption regression: token burn per unit of work jumped without any change in user behaviour or agent configuration
Source
Reported by GitHub issue, openai/codex 14593. Verified against the primary report.
A consumption regression is invisible from inside the tool and obvious from outside it: per-call cost baselines with drift alerts flag the first session that costs 5x normal for the same task class, before a fifth of the weekly allowance is gone.
The six rule types that contain this class of failure
Per-transaction cap
Any single spend above your ceiling is BLOCKED before it moves.
Daily total
Cumulative spend across all agent calls, bounded per day.
Velocity limit
Stops runaway retry loops — the #1 cause of overnight losses.
Merchant allowlist
Only approved destinations can ever receive funds.
Category rules
Flag high-risk classes (crypto, infra, refunds) for review.
Approval threshold
Above a value, the action waits for a human.
Related incidents
OpenAI confirmed Codex background tasks were silently burning users' usage limits faster than normal
OpenAI (Codex) · Jun 30, 2026
Uber burns through its entire 2026 AI coding budget in four months
Uber · Jun 2, 2026
Company hit with a reported $500 million Claude bill after leaving usage limits off
Unnamed enterprise (reported via Axios) · May 28, 2026
Don't be the next entry
Every incident in this database is the result of trusting a prompt, a provider cap, or a human review cycle. sipi.bot replaces all three with one deterministic call. 71 documented failures, one control.