Cloud Security Podcast

The Security Gaps in AWS Bedrock & Azure AI You Need to Know

Sep 23, 2025
In this insightful discussion, Kyler Middleton, a Principal Developer specializing in AI solutions, and Sai Gunaranjan, a Lead Cloud Architect for Azure, dive into the security pitfalls of AWS Bedrock and Azure AI. They highlight Azure's alarming data routing defaults that pose compliance risks and AWS’s struggle with logging and resource security. Both experts emphasize the crucial skills needed for transitioning from cloud to AI security and recommend starting with secure defaults and small-scale projects like chatbots to ensure safety in AI deployments.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
ADVICE

Treat Chatbots Like Production Services

  • Secure and monitor chatbots used across internal teams like HR and marketing.
  • Kyler advises treating chatbots as systems that need the same protection and observability as other services.
INSIGHT

Model Choice Impacts Compliance And Safety

  • Model choice affects security, compliance, and outputs beyond cost considerations.
  • Kyler warns models from certain regions can censor or refuse needed content, causing operational and patient-safety risks.
ADVICE

Use Vendor AI With A Business Impact Agreement

  • Use cloud vendor AI offerings with formal agreements for sensitive data.
  • Kyler recommends private cloud AI services with BIAs rather than public free chat tools for regulated information.
Get the Snipd Podcast app to discover more snips from this episode
Get the app