NotC2: AI-Powered C2 Framework Abuses AWS Bedrock for Stealthy Cloud Command and Control

Varonis Threat Labs disclosed NotC2, a proof-of-concept C2 framework that weaponizes Amazon Bedrock to translate plain-text instructions into AWS API calls. After obtaining valid IAM credentials, attackers execute post-compromise command and control entirely through native AWS services—including Lambda, SNS, SQS, S3, and CloudWatch—blending malicious traffic with legitimate control-plane activity. Because this is service abuse rather than an AWS vulnerability, no patch exists. Detection requires monitoring CloudTrail logs and Bedrock API behavior rather than traditional network-based C2 indicators.

• Research Overview: NotC2 Framework

  • Developed by Varonis Threat Labs as a proof-of-concept AI-powered C2 channel.
  • Leverages Amazon Bedrock for natural language to AWS API translation.
  • Discussed at Black Alps conference; analyzed by Check Point Research and Cyber Security News.

• Attack Mechanics: Cloud-Native C2

  • Requires compromised AWS credentials with sufficient IAM permissions.
  • Uses Bedrock to convert operator commands into authenticated AWS API calls.
  • Routes commands through SNS notifications, SQS queues, Lambda functions, and S3 storage.
  • Eliminates external C2 infrastructure; all traffic originates from within the AWS control plane.

• Impact and Detection Challenges

  • Activity appears as authorized administrative behavior, evading traditional network IOCs.
  • High detection difficulty due to blending with legitimate AWS management traffic.
  • Enables lateral movement and persistent access within breached AWS accounts.
  • No software patch available; classified as service abuse, not a platform vulnerability.

• Defensive Controls and Mitigation

  • Enforce phishing-resistant MFA and least-privilege IAM policies for all access keys.
  • Enable and audit AWS CloudTrail logs for Bedrock and control-plane API anomalies.
  • Monitor for unusual natural language prompts and API call sequences from Bedrock.
  • Shift detection focus from external IP/domain indicators to cloud identity and API behavior.

• Conclusion

  • NotC2 demonstrates the evolution of living-off-the-cloud tradecraft using generative AI.
  • Defenders must prioritize credential protection and cloud-native anomaly detection.
  • Traditional perimeter and network-based C2 detections are ineffective against this model.