What happened
AI Magazine reports that Hadrian, an Amsterdam-headquartered agentic AI offensive security company, announced a $40m funding round, bringing its reported total funding to $65m. The round was co-led by Forgepoint Capital International and SmartFin, with participation from existing investors. Hadrian says it will use the capital to expand in the US and EMEA and increase investment in engineering and research.
AI Magazine reports that Hadrian combines continuous exposure management with agentic penetration testing. Its Atlas product is described as continuously mapping an organization's external attack surface and using purpose-built AI agents to assess which exposures are exploitable. Nova is described as providing on-demand agentic penetration testing intended to operate at or beyond the level of a human pentest. The source does not independently verify these capabilities.
According to AI Magazine, the two products share context so security teams can move from exposure monitoring to deeper testing without beginning again. Hadrian cited results including 10 times greater visibility into critical exposures, 80% faster time to resolution, and five times the return on investment compared with manual penetration testing. These are claims attributed to the company; the article does not provide methodology, a test population, customer data, or independent validation.
Source details: aimagazine.com ↗
Why it matters
The funding supports a security platform that places AI agents directly in the offensive-security workflow, rather than using AI only for analysis or administrative tasks. Hadrian says its Atlas and Nova products can continuously assess external exposure and conduct on-demand penetration testing. If independently validated, that approach could help security teams identify and prioritize exploitable weaknesses more quickly, but the reported performance claims remain company-provided.
This is a consequential AI-security development because the reported product applies agents to an adversarial, operational task: emulating hacker attack paths to find weaknesses before attackers exploit them. That could make frequent testing more practical for enterprise teams that cannot repeatedly commission traditional penetration tests. It also raises a distinct safety requirement: systems capable of probing defenses need authorization boundaries, safeguards against unintended impact, and clear human control.
AI Magazine quotes CEO Rogier Fischer saying the company believes organizations are automating the wrong parts of offensive security. The report also describes Hadrian's leadership as combining startup and ethical-hacking experience. Those statements explain the company's positioning, not independent evidence that its agents outperform skilled security professionals or reduce real-world breach risk.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
The source does not document pricing, public availability, customer eligibility, specific underlying models, or required human oversight. Further evidence should clarify how Hadrian limits autonomous testing, how its metrics were measured, and whether customers can use the platform broadly or only through managed engagements.
The source identifies enterprise security teams as the intended users, but it does not say whether Atlas or Nova are generally available, sold as software, delivered as a managed service, or restricted by geography or customer onboarding. Pricing, deployment requirements, supported environments, and access controls are unknown.
The most useful next evidence would be independently reproducible testing, disclosed customer outcomes, and details on authorization and human review. Expansion across the US and EMEA may broaden deployment, but the report does not identify new customers, contracts, revenue, or a timetable for that expansion.