Security Feed
  1. Archives
Showing only posts by Manideep Konakandla. Show all posts.

Jul 08 2026 Designing for the inevitable: System prompt leakage and mitigations in generative AI applications

Source

System prompts form the foundation of generative AI applications. A system prompt is a collection of instructions and operational context provided to a large language model (LLM) that shapes how the model behaves and interacts with users and tools. System prompts often contain proprietary information, including role definitions, behavioral …

Posted by Manideep Konakandla on Wed 08 July 2026 in AWS Security.

Tags: Advanced (300), Amazon Bedrock Guardrails, Artificial Intelligence, Customer Solutions, Experience-Based Acceleration, Featured, Generative AI, Security, Security, Identity, & Compliance, Security Blog.

Categories

  1. Ars Technica
  2. AWS Security
  3. BleepingComputer
  4. Brian Krebs
  5. Bruce Schneier
  6. GCP Security
  7. Google Project Zero
  8. The Daily Swig
  9. The Guardian
  10. The Register
  11. Threatpost

Tag cloud

  • Security
  • Uncategorized
  • Security, Identity, & Compliance
  • Security Blog
  • Microsoft
  • Biz & IT
  • AI
  • Security & Identity
  • CryptoCurrency
  • Artificial Intelligence
  • google
  • Announcements
  • Legal
  • Foundational (100)
  • A Little Sunshine
  • privacy
  • Apple
  • Intermediate (200)
  • Mobile
  • hacking
  • squid
  • Advanced (300)
  • LLM
  • Technical How-to
  • vulnerabilities

Security Feed. Powered by Pelican and m.css. Code is available on GitLab.