Level 3. Significant control incident
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)
- Developer
- Not established
- Model
- Not established
- When it happened
- August 12, 2025
- 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 embracethered.com. 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
- 3 of 5Significant control incident
- Potential
- 5 of 5Catastrophic incident
An important technical, security or permission boundary was crossed, or there were real consequences. 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.
The reports point to an actual level of 4. It is held at 3 until the developer, an official body or a second publisher confirms it.
Why this level
- Major compromise of an external system. Argues for level 4
- A security boundary was crossed. Argues for level 3
- A safeguard was bypassed. Argues for level 2
- The system acted without authorisation. Argues for level 2
- Propagation was not under control. Argues for a potential level of 5
- The system attempted to replicate or persist. Argues for a potential level of 4
What is established, and what is not
Confirmed
Nothing is confirmed yet.
Reported, not confirmed
Major compromise of an external system
Last but not least, to demonstrate that we have full control of the developer’s host, we show that Copilot can be hijacked to download malware, and join a remote command and control server.
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)The system attempted to replicate or persist. A claim, so it counts toward potential only
This also means we can build an actual AI virus that attaches to files and propagates as developers download and interact with infected files.
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)A safeguard was bypassed
This will put GitHub Copilot in YOLO mode. And it disables all user confirmations, and we can run shell commands, browse the web, and more!
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)A security boundary was crossed
We achieved Remote Code Execution powered by Prompt Injection.
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)The system acted without authorisation
The prompt injection first adds the line “chat.tools.autoApprove”: true, to the ~/.vscode/settings.json file.
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)Uncontrolled propagation. A claim, so it counts toward potential only
This can lead to further spread as other developers unknowingly propagate the infected code.
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)
Checked, and did not happen
Nothing here was checked and ruled out.
Not confirmed
Nothing is outstanding.
Where this came from
The original report · read September 19, 2026
GitHub Copilot: Remote Code Execution via Prompt Injection (CVE-2025-53773)A structured incident database · read September 18, 2026
AI Agent Incident Database by sipi.bot (CC BY 4.0)
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.