Protect agentic AI applications with Datadog AI Guard

The Datadog team discovered and mitigated a malicious attempt to inject a backdoor into their open-source repositories via a seemingly helpful AI-generated contribution. An attacker used an AI agent to create a pull request with subtle code changes designed to exfiltrate sensitive data, but Datadog's rigorous review process and automated checks flagged the suspicious activity. This incident highlights the growing risk of AI-assisted malicious contributions to open-source projects and the importance of strong security practices.