Escape raises $18M to replace pen-testers with AI agents for offensive security testing

Escape Secures $18 Million Series A Funding for AI-Powered Security Platform

Cybersecurity startup Escape has closed an $18 million Series A funding round led by Balderton Capital, with participation from Uncorrelated Ventures and existing backers IRIS and Y Combinator1

1

. The investment brings the company's total funding to $23 million since its founding in 2020 and arrives at a moment when traditional security approaches struggle to keep pace with AI-generated code1

1

. Founded by Tristan Kalos and Antoine Carossio, two French engineers who met at UC Berkeley, Escape has built what it calls an offensive security engineering platform designed to identify vulnerabilities that emerge only in live production environments2

2

.

AI Agents Replace Manual Penetration Testing at Scale

The core of Escape's approach centers on AI agents that simulate attacker behavior against live systems rather than waiting for vulnerabilities to surface after deployment2

2

. These agents continuously map attack surfaces, interact with authentication flows and business logic, then generate proof-of-exploitation to demonstrate exactly how flaws can be triggered1

1

. The AI agent platform also provides remediation guidance and reproduction steps so security teams can verify patches haven't introduced new problems. This shift from reactive scanning to active testing addresses what Balderton partner Suranga Chandratillake described as an impossible dilemma: "Rely on legacy scanners, knowing they do not have the quality of pen-testing, or work with manual offensive security teams and fail to scale to the volume of code"2

2

.

Research Exposes Thousands of Vulnerabilities in AI-Built Applications

The timing of this Series A funding follows research published by Escape in October 2025 that scanned more than 5,600 publicly deployed applications built with vibe coding platforms like Lovable, Base44, and Bolt.new1

1

. The results revealed over 2,000 high-impact vulnerabilities, hundreds of exposed secrets, and cases of personal data sitting exposed in live production systems accessible to anyone who knew where to look1

1

. This research highlighted a critical gap: traditional security tools were built for a world where code was written slowly and reviewed carefully, but that world no longer exists as AI-generated code proliferates.

Platform Integration and Growing Customer Base

Escape integrates into engineering workflows via CI/CD pipelines, allowing vulnerabilities to surface before code reaches users1

1

. The platform performs continuous attack-surface discovery and executes testing routines against identified components to identify vulnerabilities in application logic, configuration, and integrations after code is deployed2

2

. Typical use cases include testing APIs, validating multitenant application security, identifying exposure of sensitive data, and evaluating business logic vulnerabilities such as insecure direct object references2

2

. The company claims more than 2,000 security teams globally now use its platform to run over 300,000 security assessments monthly, with notable customers including BetterHelp, PandaDoc, CyberCube Analytics, and Arkose Labs2

2

. Escape has also achieved month-on-month revenue growth of 15% or more1

1

.

Expansion Plans and Team Diversity

The new funding will be deployed to develop agentic penetration testing capabilities and roughly double the 32-person team over the coming year while expanding go-to-market operations across the US and Europe1

1

2

2

. Kalos emphasized that security teams are outnumbered and managing siloed, manual processes in a world where code is written and attacked at the speed of AI1

1

. The team already reflects an unusual degree of diversity for a cybersecurity startup, with 30% female representation and more than 12 nationalities, which the company intends to maintain as it scales1

1

. As both human and agentic developers multiply, the emphasis on live environments rather than code repositories positions Escape to address security risks that only emerge when configurations and authentication flows run in production.