Cloud Vulnerabilities Jump Fourfold in Months, and Old Access Keys Add to the Risk
Datadog has claimed that the median number of high or critical vulnerabilities in cloud workloads increased fourfold between May and November this year. Its State of Cloud Security 2026 report, published on November 10, links the rise to the impact of frontier AI. The company says the median went from four in May, when Anthropic's Mythos model was released, to 14 in June and 16 in November.
The findings are based on thousands of customer organisations worldwide. The vulnerability data comes from third-party libraries used by applications monitored through Datadog Code Security's Software Composition Analysis feature. Datadog says this shows the growing number of entries in the CVE database is not a theoretical risk, because it affects a rising number of real applications running in the cloud.
The report also points to long-lived credentials, especially for cloud-hosted AI inference services and CI/CD pipelines. Datadog warns these are consistently among the most common causes of cloud breaches, and that automated tools keep scanning for over-permissioned credentials and exposed keys. It says autonomous attackers could use the same tooling. Nearly half of organisations using Gemini rely on long-lived Gemini API keys, and almost 30% of AWS IAM and Google Cloud Service Account users have access keys older than three years.