Fleuret AI raises a €4 million pre-seed round led by RAISE Ventures for agentic AI penetration testing

Fleuret AI raises a €4 million pre-seed round led by RAISE Ventures for agentic AI penetration testing

5 October 2026 — Fleuret AI, a Paris startup that automates penetration testing with AI agents, announced on 5 October 2026 that it has closed a €4 million pre-seed round. The company wants to turn penetration testing from a one-off audit into a continuous process.

Not yet independently verified. The round rests on the company's own announcement of 5 October; the only other coverage found is one European startup outlet. Customer names come from the company. We will update this when it can be confirmed, and remove this note.

The round

According to Fleuret's announcement, the €4 million pre-seed round is led by RAISE Ventures, alongside Auriga Cyber Ventures, Wind Capital and Better Angle, and business angels from the European cybersecurity sector. The company says the money will go into hiring in AI, software engineering and offensive security, and into building out its platform.

What the company does

Fleuret's platform uses AI agents that map an organisation's environment, explore its web applications, APIs and infrastructure, look for vulnerabilities and attempt to exploit them to confirm they are real, then re-test as systems change. The company positions this as a European alternative running on European infrastructure, aimed at organisations with sovereignty concerns and NIS2 obligations. It was founded by Yanis Grigy and Augustin Ponsin, and its announcement names customers in insurance, marketing software and bug bounty.

Why it matters for buyers

Traditional penetration tests are expensive and happen once or twice a year, while code and infrastructure change every week. AI agents that test continuously promise broader and more frequent coverage. They also raise new questions: an automated tool that actively exploits systems needs tight scoping, safeguards against damage and clear evidence of what it did.

What to check

• How targets and scope are defined and enforced, and how the agents avoid disrupting production systems.
• What evidence each finding comes with, and how false positives are handled.
• Whether results satisfy your auditors, insurers or regulators in place of a human-led test.
• Where testing traffic originates and where findings are stored.
• How the service complements, rather than replaces, periodic human-led testing for complex logic flaws.

Company profile on TrustList: Fleuret AI

Sources

• Fleuret AI: Fleuret raises €4M — 5 October 2026
• Startup Rise EU: Fleuret AI raises €4M — 5 October 2026