TLDR: Synack has launched Sara (Synack Autonomous Red Agent), an agentic AI architecture designed to transform Penetration Testing as a Service (PTaaS). This new platform integrates autonomous AI capabilities with human expertise, offering proactive, risk-based security validation to combat the rising threat of AI-powered cyberattacks. Key components include Sara Triage for autonomous vulnerability validation and Sara Pentest for comprehensive, objective-based penetration tests.
REDWOOD CITY, Calif. – August 11, 2025 – Synack, a leader in offensive security innovation, today announced the launch of its groundbreaking agentic AI architecture, Sara (Synack Autonomous Red Agent). This new platform is set to redefine Penetration Testing as a Service (PTaaS) by seamlessly integrating advanced AI capabilities with the critical oversight of human experts, a ‘human-in-the-loop’ approach.
Sara is designed to provide proactive, risk-based security validation, directly addressing the escalating threat landscape where adversaries increasingly leverage AI to develop, optimize, and scale their attacks. The platform embodies an ‘AI-versus-AI’ model, where AI-powered validation, supervised and guided by human judgment from the Synack Red Team, actively counters machine-driven reconnaissance and attacks. This hybrid intelligence approach aims to deliver a powerful, adaptive solution that mirrors real-world adversary behavior while significantly minimizing risk and false positives.
Dr. Mark Kuhr, Synack co-founder and CTO, emphasized the urgency of this innovation, stating, “Security teams are no longer just fighting humans—they’re defending against AI-enhanced adversaries. With our Sara agentic AI and human-in-the-loop model, we equip defenders with the same level of intelligence and speed, turning the tables in this era of AI-versus-AI cybersecurity.”
The Sara agentic AI architecture is engineered to deliver scalable and adaptable assessment of attack surface risk. The first core component, Sara Triage, is available immediately. It provides autonomous triage of discovered vulnerabilities, ensuring that only truly exploitable findings are validated. Later this year, Synack plans to release Sara Pentest, which will conduct full-scope, objective-based penetration tests in collaboration with the Synack Red Team.
Also Read:
- Agentic AI Revolutionizes Cybersecurity Response, Demonstrating Significant Time Reductions at Black Hat USA 2025
- KnowBe4 Bolsters Cybersecurity Defenses with Expanded Agentic AI Capabilities
Synack highlights that Sara’s human-in-the-loop architecture is crucial for the discovery of complex logic flaws, chained exploits, and nuanced vulnerabilities that often elude purely automated systems. This integration bridges the gap between rapid automated detection and the invaluable intuition and critical thinking of human security researchers. The company, a pioneer in human-led and AI-powered PTaaS, is committed to leveraging agentic AI innovations and its vetted community of security researchers to provide continuous penetration testing and autonomous vulnerability management, ultimately enhancing global security.
