← 返回列表
AI and Cloud Security Research
单站
Fog Security Blog
入库 2026-11-18T05:50:13
AI and Cloud Security Research
November 5, 2026Google Gemini's Safety Filter Refused. The AI Agent Delivered Anyway.The model said no. The download link said yes. How Google Gemini served images its own safety guardrails had refused.Read the research →

All posts
May 12, 2026Authorization Bypass in Amazon Quick: Unauthorized AI Chat Agent UsageWe discovered an authorization bypass in Amazon Quick’s AI Chat Agents that allows users to access and interact with AI agents despite explicit administrative restrictions. AWS responded by deploying a fix without notifying customers, classified the issue as “none,” and did not publish an advisory.

April 27, 2026Disabling SSE-C Encryption For Amazon S3 Buckets: What to Know for Data Protection and Ransomware PreventionA technical analysis of a new AWS change to disabling SSE-C encryption. Read more about how we interpret AWS's design decisions and how we can improve our security by to protect against unique S3 ransomware.

April 16, 2026State of Bug Bounties with AI: Analysis of curl and other ProgramsReview of AI impact on bug bounty programs in 2026 including triage impact, report volume impact, and more. Analysis done on curl and other HackerOne programs.

August 20, 2025Mistrusted Advisor: Evading Detection with Public S3 Buckets and Potential Data Exfiltration in AWSWe found security issues with AWS's Trusted Advisor Service that allowed us to create public S3 buckets while evading detection for potential data exfiltration.

August 5, 2025Introducing aws-size: A New Open Source Tool for Difficult Service Limit Visibility in AWSReleasing aws-size, a new open source tool to help with managing difficult service limits such as IAM policy size, bucket policy size, SCPs, RCPs, EC2 user data, and more. These limits may not be covered by AWS services such as Service Quotas or Trusted Advisor.
.png)
July 15, 2025IAM SAR-ry: Deep Dive and Nuances of AWS's Programmatic IAM Action List and Service Authorization References (SAR)AWS released a programmatic service reference for IAM. We did research and pulled statistics such as number of IAM actions, inconsistencies with the service authorization pages, access levels, changes in permission classification, and how this impacts your cloud security and IAM workflows.
%20copy.png)