#ai-code-execution

[ follow ]
Information security
fromThe Hacker News
3 hours ago

AI Flaws in Amazon Bedrock, LangSmith, and SGLang Enable Data Exfiltration and RCE

Amazon Bedrock AgentCore Code Interpreter's sandbox allows outbound DNS queries, enabling attackers to exfiltrate data and establish command-and-control channels despite network isolation configuration.
[ Load more ]