AWS launches frontier agents for security testing and cloud operations
New autonomous AI agents cut penetration testing by 90% and resolve incidents 3-5x faster.
AWS has officially launched its new class of autonomous AI systems called 'frontier agents,' with the AWS Security Agent and AWS DevOps Agent now generally available. Unlike traditional AI assistants that handle single tasks, these agents operate independently for hours or days to achieve complex goals. The Security Agent transforms penetration testing from a periodic, costly bottleneck into a 24/7 on-demand capability. By ingesting source code and architecture diagrams, it identifies and exploits vulnerabilities, finding attack chains traditional scanners miss. Preview customers like Bamboo Health and HENNGE K.K. report reducing testing duration by over 90%, compressing timelines from weeks to hours.
The AWS DevOps Agent acts as an autonomous operations teammate across AWS, Azure, hybrid, and on-premises environments. It integrates with observability tools like Datadog and Splunk, and code repositories like GitHub, to investigate incidents and optimize reliability. In preview, it achieved metrics including 75% lower Mean Time to Resolution (MTTR), 80% faster investigations, and 94% root cause accuracy, supporting 3–5x faster incident resolution overall. These frontier agents represent a shift from tools requiring constant human direction to intelligent systems that understand context, reason through problems, and take persistent action, fundamentally changing how organizations approach application security and cloud operations.
- AWS Security Agent autonomously performs penetration testing, reducing timelines from weeks to hours (over 90% faster) by identifying complex attack chains.
- AWS DevOps Agent resolves incidents 3–5x faster, with preview results showing 75% lower MTTR, 80% faster investigations, and 94% root cause accuracy.
- Both are 'frontier agents'—autonomous systems that scale massively, run persistently without constant oversight, and work across multicloud and on-prem environments.
Why It Matters
This automates high-skill, time-intensive security and ops work, enabling continuous protection and faster incident resolution at cloud scale.