Terra Security
Agentic AI penetration testing that continuously tests web apps with human oversight
Last reviewed 2026-06-19
Terra Security is an agentic AI penetration-testing platform for continuous web application security. Instead of point-in-time manual pentests, Terra deploys a coordinated swarm of AI agents that scope the environment, discover attack surfaces, generate attack hypotheses, and validate vulnerabilities, building a custom test plan per customer and re-running tests automatically when new vulnerabilities or changes appear. Each application is tested by dozens of agents tailored to the organization's profile. Crucially, Terra is human-in-the-loop by design: a human security expert supervises the agent swarm, and when agents hit limits a gateway lets human testers operate inside the same agentic workflow. This keeps accuracy and reliability high and means it operates as a supervised agent rather than a fully autonomous attacker. Founded in 2024 by Shahar Peled (CEO) and Gal Malachi (CTO), headquartered in New York with a team in Israel, Terra raised an $8M seed and a $30M Series A led by Felicis, for a reported $38M total.
What it can do
Continuously pentest web applications
SupervisedA swarm of AI agents runs thousands of advanced tests tailored to each organization, re-testing automatically when new vulnerabilities or changes are discovered.
sourceDiscover attack surface and validate vulnerabilities
SupervisedCoordinated agents scope environments, discover attack surfaces, generate hypotheses, and validate vulnerabilities rather than just flagging potential issues.
sourceHand off to human testers via a gateway
SupervisedWhen agents encounter limits, a gateway lets human security experts operate within the same agentic workflow, keeping a human supervising the swarm.
source
Strengths
- +Continuous, agent-driven pentesting replaces slow point-in-time manual tests
- +Validates vulnerabilities rather than just listing potential issues, reducing false positives
- +Human-in-the-loop gateway keeps expert oversight on offensive actions
Limitations
- −Offensive testing is high-stakes, so human supervision is required, not optional
- −Focused on web applications today (network/red-teaming expansion planned)
- −Enterprise-only with no public pricing
Overview
Terra Security is an agentic AI penetration-testing platform for continuous web application security. A coordinated swarm of AI agents, supervised by a human expert, replaces slow point-in-time manual pentests.
What it does
Agents scope the environment, discover attack surfaces, generate attack hypotheses, and validate vulnerabilities, building a custom test plan per customer and re-testing automatically when the app changes or new vulnerabilities appear. A gateway lets human testers step into the same agentic workflow when agents hit limits, keeping a human in the loop, so Terra operates as a supervised agent.
Integrations & setup
Fits into security and development workflows (e.g., CI/CD, ticketing). Sales-led enterprise deployment.
Pricing
Enterprise, sales-led; no public pricing.
Best for / not for
Best for security teams that want continuous, validated web-app pentesting with expert oversight. Less suited to teams wanting a free scanner or a fully hands-off autonomous attacker. Network coverage and AI red teaming are planned expansions.
Traction
Founded in 2024 by Shahar Peled (CEO) and Gal Malachi (CTO), headquartered in New York with a team in Israel. Raised an $8M seed (SYN Ventures, Underscore VC) and a $30M Series A led by Felicis with Dell Technologies Capital, for a reported $38M total.
Alternatives
XBOW is the closest autonomous offensive-security competitor; Prophet Security and Dropzone AI cover defensive SOC automation.
What people are saying
We aggregate real LinkedIn discussion into sentiment for the agents people search most. Terra Security isn't tracked yet, want it added? Request tracking.
FAQ
Is Terra a fully autonomous hacker?+
No. Terra runs a swarm of AI agents that scope, discover, and validate vulnerabilities, but a human security expert supervises them and a gateway hands off to human testers when agents hit limits. It is a supervised agent by design, for accuracy and safety.
How is it different from a traditional pentest?+
It is continuous rather than point-in-time: agents build a custom test plan per app and re-run tests automatically as the application changes or new vulnerabilities emerge.
Sources
- Terra Security raises $8M for agentic AI pentesting (SecurityWeek) · accessed 2026-06-19
- Terra Security raises $30M Series A (Terra blog) · accessed 2026-06-19
- Terra Security raises $30M for AI pentesting (Calcalist) · accessed 2026-06-19
Last reviewed 2026-06-19