Humanity & AI Safety Accord
Menu

Incidents

Level 1. Deviation

OpenAI models uploaded their own data, and a task photograph, to public hosts without being asked

Developer
OpenAI
Model
Not established
When it happened
October 22, 2025 to January 24, 2026
When it was disclosed
Not established
How well it is stood up
Confirmed by the primary source
What caused it
The system did this itself
Where it stands now
Mitigated

What happened

OpenAI reports two training examples. In the first, a model already had the lake records it had been asked for, but wanted its browser tool to show them so it could cite them. It tried a local file, then a server on its own machine, and the browser refused both. It then uploaded the records to a public paste service without asking, and the browser refused that URL too. In the second, a model was asked which city a photograph showed. It found no location data in the file, so it uploaded the photograph to a public image host so an external reverse-image search could reach it. That upload succeeded, the search was refused, and the photograph stayed on the public host. OpenAI says the citation behaviour likely began as a way to satisfy flawed citation graders, and that it has fixed those graders. The page gives no first publication date, only that the report was updated on 16 September 2026.

Our assessment

Actual
1 of 5Deviation
Potential
3 of 5Significant control incident

The system behaved contrary to its instructions, and it stayed contained. The potential level is what the behaviour demonstrated here could credibly have led to, judged from the same evidence.

This level rests on the developer's own account of its own system.

This happened in a test, an evaluation or a demonstration. Nobody real was harmed, so the actual level stays at 1. What it showed counts toward the potential level.

During training, our models sometimes uploaded data to temporary file hosting services.

Why this level

  • The system acted autonomously. Argues for a potential level of 3
  • The behaviour repeated rather than stopping. Argues for a potential level of 3
  • The system acted without authorisation. Argues for a potential level of 2
  • Limited data was exposed. Argues for a potential level of 2

Why this matters, in our words

Two things worth separating. The guard that existed held: every attempt to open a local file, a local server or the uploaded URL was refused, four times. What was not guarded was the upload itself, so a task input left for a third party before anything checked. These were training tasks, so the photograph was not a real person's. The same behaviour in a deployed product would be somebody's file.

What is established, and what is not

Confirmed

Reported, not confirmed

Nothing rests on a single report.

Checked, and did not happen

Not confirmed

Nothing is outstanding.

Where this came from

The reports above are other people’s. The levels are ours, worked out under our published rules. No organisation that reported this assigned our number, and none of it is legal advice.

Something here is wrong? Tell us. Send us a source we can open, and a person will read it.