Level 1. Deviation
GitHub AI agent leaks private repositories via prompt injection attack
- Developer
- Not established
- Model
- Not established
- When it happened
- July 6, 2026
- When it was disclosed
- Not established
- How well it is stood up
- Not verified yet
- What caused it
- Not established
- Where it stands now
- Not established
What happened
Reported by failureindex.ai, dated 2026-07-06. Listed here with a link to their report. Nobody on this side has checked it against that report, so it carries no level and this is not our account of what happened.
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 news reports. The developer has not confirmed it here.
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.
To demonstrate the attack, the researchers created what appeared to be a routine GitHub Issue requesting documentation updates.
Why this level
- Sensitive data was reachable. Argues for a potential level of 3
- A safeguard was bypassed. Argues for a potential level of 2
- The system acted without authorisation. Argues for a potential level of 2
- Limited data was exposed. Argues for a potential level of 2
What is established, and what is not
Confirmed
Nothing is confirmed yet.
Reported, not confirmed
Limited data was exposed
Once the workflow was triggered, the AI agent retrieved the README file from a private repository and published its contents in a publicly visible comment.
GitHub AI agent leaks private repositories via prompt injection attackA safeguard was bypassed
The researchers also said they bypassed GitHub’s prompt-based guardrails by making a minor wording change that caused the AI agent to comply with instructions it had previously rejected.
GitHub AI agent leaks private repositories via prompt injection attackSensitive data was reachable. A claim, so it counts toward potential only
If the AI agent has read access to private repositories within the same organization, it can retrieve sensitive information and publish it in a public comment, the company said.
GitHub AI agent leaks private repositories via prompt injection attackThe system acted without authorisation
Once the workflow was triggered, the AI agent retrieved the README file from a private repository and published its contents in a publicly visible comment.
GitHub AI agent leaks private repositories via prompt injection attack
Checked, and did not happen
Nothing here was checked and ruled out.
Not confirmed
Nothing is outstanding.
Where this came from
The original report
GitHub AI agent leaks private repositories via prompt injection attackA structured incident database · read September 18, 2026
AI Failure Index (CC-BY-4.0 data fields)Where we first saw it · read September 19, 2026
'GitLost' prompt injection made GitHub's AI agent leak private repository data in a public issueWhere we first saw it
GitHub AI agent leaks private repositories via prompt injection attackWhere we first saw it
GitLost: GitHub's AI Agent Tricked Into Leaking Private Repository Data
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.