Threat Briefing: Why Attackers Target AWS Bedrock

In this episode of Threat Briefing, Lucie sits down with Vectra AI security researcher Alex Groyz to examine how adversaries are targeting cloud-hosted GenAI platforms—focusing specifically on AWS Bedrock.

Alex kicks off by outlining the cloud AI threat surface and dives into why attackers are shifting away from public or open-source models in favor of enterprise-grade platforms like Bedrock. From there, he explores how attackers gain access to inference APIs, what LLMjacking looks like in practice, and why traditional monitoring tools often miss the signals of GenAI abuse.