InfoSecBulletin Cybersecurity for mankind
A major security issue happened at Elon Musk’s AI company, xAI, when a developer accidentally posted a private API key on GitHub, making it publicly accessible for nearly two months.
Exposed credentials allowed unauthorized access to private large language models (LLMs) fine-tuned for SpaceX, Tesla, and Twitter/X, revealing serious security vulnerabilities at major tech companies.
Citrix NetScaler SAML 0-Day Flaw Under Attack
Major Danish university breached, 200,000 users at risk
Microsoft’s X account hijacked to promote Clippy crypto scam
Critical cPanel, GitLab AI Gateway and Dell CSM Flaws Enable RCE And Admin Hijacking
Nearly 100,000 email addresses exposed in first AI-related data breach in Singapore
Hackers Exploit Zimbra Mail Servers: TeamViewer patched 5 critical flaws
Google Warns of Hackers Actively Exploiting Citrix 0-Day Flaws
CISA Warns Critical MikroTik RouterOS Flaw While Cisco SD-WAN Zero-Day Exploited in Attacks
Apple Zero-Day Exploited: Pentagon Data Breach Reportedly Exposes Sensitive Data of 3 Million People
JadePuffer Agentic AI targets and destroys Azure’s cloud resources
Philippe Caturegli, Chief Hacking Officer at Seralys, discovered the security incident and shared the leak on LinkedIn.
GitGuardian, a company that detects exposed secrets in code repositories, investigated the issue.
Their automated scanning systems monitor GitHub for sensitive credentials and detected a compromised API key on March 2, 2025.
KrebsOnSecurity reports that the API key leaked by an xAI employee stayed active until April 30, despite the employee being notified right away. GitGuardian escalated the issue to xAI’s security team at that time.
According to Eric Fourrier from GitGuardian, the company found that “the key had access to at least 60 distinct data sets” including several unreleased and developmental versions of Grok models.
The leaked API credentials could have been used to query private custom-trained models including “grok-2.5V” (unreleased), “research-grok-2p5v-1018” (development), and “grok-spacex-2024-11-04” (private).
These models seem to be customized for managing internal data from Musk’s corporate network.
“The associated account not only has access to public Grok models but also to what appears to be unreleased, development, and private models,” GitGuardian noted in their communication to xAI.
The leak happened when a .env file was accidentally added to a public GitHub repository.
xAI did not provide public comment on the incident, and the repository containing the key was removed shortly after GitGuardian’s notification to the security team.
Caturegli noted that such “long-lived credential exposure highlights weak key management and insufficient internal monitoring, raising questions about safeguards around developer access and broader operational security”.
