A Reset Button Mention Puts Codex and ChatGPT Work on Early Watch

Tibo Sottiaux posted a short, suggestive message on August 27: “A good thing about having aged is that I feel that it’s been 20 years since I’ve pressed the reset button. Intrigued to see if I can find it tomorrow and dust it up”. The wording is enough to put Codex and ChatGPT Work users on early watch for a possible usage-reset development.

It is not enough to call a reset confirmed. The post does not identify the product surface, eligible plans, account group, reset mechanism, or exact time. “Tomorrow” is also relative to the post and does not establish a universal rollout window or timezone.

What the signal may mean

The most relevant interpretation is that OpenAI could be preparing to reset, refresh, or otherwise change access to usage connected with Codex and ChatGPT Work. That matters to developers who plan larger coding-agent runs around available capacity: a reset could affect when queued work, repository migrations, or intensive debugging sessions become practical again.

That interpretation remains an inference from an individual public post. It is not an OpenAI release note, plan announcement, or account-level entitlement notice. Readers should treat it as a watch signal, not as permission to delay critical work or assume that a quota will return.

How OpenAI’s plan documentation fits

OpenAI’s help documentation explains Codex usage in the context of ChatGPT plans. That official context is useful because Codex access and limits can depend on the plan and the way the product is used. It also means that a public hint may not translate into the same result for every subscriber, workspace, or account.

The documentation should be the reference point for any confirmed change. Until it is updated or an in-product notice appears, the August 27 post does not prove that all users will receive a reset, that the reset will happen on August 28, or that it will apply to ChatGPT Work specifically.

The banked-reset question is still open

Nothing in the post says that a possible reset would be a banked reset. A banked reset has different planning implications from a regular replenishment: users would need to know whether unused capacity can accumulate, how much can be carried, when it expires, and which plans qualify. Those details are absent here.

The same caution applies to audience. The post does not say whether the signal concerns individual ChatGPT users, ChatGPT Work workspaces, Codex users across multiple plans, a test cohort, or an internal demonstration.

What to watch next

The useful confirmation triggers are concrete: an OpenAI Help update, an in-product usage change, an official announcement that names Codex or ChatGPT Work, or account-level behavior showing a reset. Until one of those appears, keep normal capacity plans in place and avoid presenting the post as a promised benefit.

For now, the correct headline is simple: a possible reset may be worth watching, but the reset itself, its timing, its audience, and any banked behavior remain unconfirmed.

Sources and methodology

This early-watch note uses Tibo Sottiaux’s public X post as the signal and OpenAI’s official Codex plan documentation for product context. No additional rumor or secondary reporting is used. Claims about a possible reset are explicitly framed as inference; no reset is treated as confirmed.