Fleuret AI raises €4M for agentic pentesting

French startup Fleuret AI raised €4 million in pre-seed funding to build continuous penetration-testing agents, with its announcement describing tools that map systems, test vulnerabilities, and verify remediation. The company says its platform watches changes to an attack surface, turning a one-off audit into repeat testing; buyers should treat its findings as leads for human security review, not autonomous authorization. This is the permission boundary in AISI’s agent-evaluation findings: a tool that can probe production systems needs explicit scope and oversight.