GitGuardian highlights the urgent need for enterprises to map their expanding "credential layer" as rapid software production and AI agent adoption increase the risk of exposed secrets.
Key Points
- GitGuardian reports that 28.65 million new hardcoded secrets were detected in public GitHub commits in 2025, a 34% year-over-year increase.
- Research indicates that 22% of initial breach access stems from compromised credentials, often found on unmanaged developer endpoints or collaboration systems.
- Infostealer malware, such as Shai-Hulud and S1ingularity, frequently targets developer laptops, where researchers found an average of nearly five secrets per compromised machine.
- AI coding agents introduce new security vulnerabilities, with over 2,000 valid secrets identified in public Model Context Protocol configuration files during 2025.
- Effective security requires a "Detect, Remediate, and Prevent" strategy that inventories credentials across repositories, endpoints, and cloud environments to provide necessary context.