Member-only story
Discovering publicly exposed data in AWS S3 and Google Cloud Storage
Open cloud storage buckets continue to be a major source of data breaches. In this article, I introduce Cloudhound, a powerful Go tool for discovering and analyzing publicly accessible cloud storage buckets across AWS and Google Cloud. The tool uses advanced techniques to find exposed data, efficiently deduplicates content, and prioritizes potentially sensitive information.

The Hidden World of Public Buckets
Cloud storage has revolutionized how businesses store data, but misconfiguration remains rampant. A staggering number of organizations inadvertently expose sensitive information through improperly secured S3 buckets and Google Cloud Storage containers. Security researchers have discovered everything from personal customer data to corporate secrets, API keys, and database backups sitting in open buckets.
While companies like Amazon and Google have improved their default security settings, the problem persists — often due to human error, legacy systems, or a simple lack of understanding about cloud security models.
This article presents Cloudhound, a sophisticated tool written in Go that helps security researchers and organizations identify publicly accessible buckets to assess exposure risk.
How Cloudhound Works
Cloudhound takes cloud bucket discovery to the next level through several key strategies:
- Multi-faceted discovery — Uses permutations, common patterns, company names, and dictionary-based approaches
- Global coverage — Searches across both AWS S3 and Google Cloud Storage platforms
- Intelligent content prioritization — Ranks discovered files by potential sensitivity
- Database-backed deduplication — Avoids redundant downloads using content hashing
- Optimized performance — Employs concurrent scanning with configurable parallelism
The tool is designed to be ethical by default. It only accesses buckets that are intentionally configured to allow public access, and it respects rate limits to avoid overloading services.