OpenAI Addresses Security Incident Involving Third-Party Library
OpenAI has reported a security incident involving a compromised third-party developer library. The company stated it found no evidence of user data access, system compromise, or software alteration.
On March 31, OpenAI identified a security issue involving Axios, a widely used third-party developer library. The company stated the incident was part of a broader software supply chain attack attributed to actors believed to be linked to North Korea.
What Happened: Technical Details
The attack targeted an automated GitHub Actions workflow used by OpenAI, causing it to download and execute a version of Axios that contained malicious code.
This automated workflow had access to a certificate and notarization material used for digitally signing several macOS applications to verify their legitimacy. The affected applications included:
- ChatGPT Desktop
- Codex
- Codex-cli
- Atlas
OpenAI's Analysis and Findings
Following an internal investigation, OpenAI's analysis concluded the following key points:
- The signing certificate present in the workflow was likely not successfully exfiltrated by the malicious payload.
- There is no evidence that user data was accessed, company systems or intellectual property were compromised, or that any OpenAI software was altered.
- Passwords and OpenAI API keys were not affected by this incident.
- The root cause was identified as a misconfiguration in the GitHub Actions workflow. OpenAI stated this misconfiguration has been addressed.
Response and Required User Actions
In response to the incident, OpenAI is taking several steps:
- Updating its security certifications.
- Requiring all macOS users to update their OpenAI applications to the latest available versions.
Effective May 8, older versions of OpenAI's macOS desktop apps will no longer receive updates or support and may cease to function.