One prompt let researchers take over every AWS AgentCore agent in a region
Zenity Labs says AgentCore’s default role let it read private chats, copy source code and steal secrets across a whole AWS account. AWS has since cut those permissions.
Via TNW | Security
Researchers at Zenity Labs used a single prompt to one public-facing AI agent on Amazon Bedrock AgentCore to take over every AgentCore agent in the same AWS account and region. The security firm published the research on Thursday, alongside a talk at the SecTor 2026 conference in Toronto.
AgentCore is AWS’s managed service for building and running AI agents. Zenity says the chain of flaws, which it calls AgentCorruption, let its researchers read private conversations, copy agents’ source code, steal…
Courtesy
This story was originally published by TNW | Security. All rights belong to the original publisher.
Read the original on thenextweb.com ↗