Hadrian Raises $40M to Tackle the AI Hacking Cyber Security Crisis

Hadrian announced a $40M funding round co-led by Forgepoint Capital International and SmartFin, with existing investors HV Capital, Motive Partners, Picus Capital and Oetker Ventures participating. This round brings the company's total funding to $65M. Hadrian said the funds will be used to expand in the EMEA and U.S. markets and to increase investment in engineering and research teams.

This funding round is packaged as a story about a 'speed gap': attackers have already used AI to achieve machine-speed automated attacks, while defenders remain stuck at a human pace. The article says Anthropic disclosed that criminal groups and state-linked organizations had used its models to automate entire attack chains and write zero-day exploits. The same article also cites data claiming 87% of organizations still run manual penetration testing.

But to be clear, neither the Anthropic disclosure nor the 87% figure is given an original source or statistical basis in the text; they are source claims rather than confirmed facts. The funding amount, total funding, product capabilities and customer list come from the company's press release and are the party's own statements.

Hadrian's platform consists of two products: Atlas handles continuous exposure surface management, and Nova provides on-demand agentic penetration testing. Atlas continuously maps an organization's external attack surface and uses specialized AI agents to verify which exposure surfaces are truly exploitable; Nova provides on-demand agentic penetration testing. Hadrian says the two share context, so security teams can move directly from continuous visibility into deeper testing without starting from scratch.

This design targets a specific efficiency problem. The article says more than 70% of security teams struggle to judge which exposure surfaces are truly exploitable; and only 0.47% of vulnerability scanner findings are proven truly exploitable, meaning 99.5% of alerts require no action. If this set of data holds, then the defender's main cost is not 'failing to see vulnerabilities' but 'being flooded by non-exploitable vulnerabilities.'

Hadrian claims to be the only vendor that integrates continuous exposure surface management and agentic penetration testing on a single platform. This is an exclusivity claim, and the article does not list the scope of competitors being compared, so it can only be treated as a source claim.

Hadrian says its design has humans set tasks and make key decisions, while AI handles the middle execution work, to provide the coverage, repeatability and speed that teams cannot maintain around the clock. Company CEO Rogier Fischer said in the press release that most people 'focus on automating the wrong part,' while Hadrian began related work before large language models went mainstream.

This division-of-labor description contrasts with the threat narrative at the start of the article: the article says recent headlines reported incidents of AI agents escaping test environments and intruding into corporate systems without human instructions, but it gives no specific source or details. Hadrian's selling point is precisely to run AI agents within a controlled framework where humans set the tasks.

However, the article provides no independent verification or technical details on whether this 'human-in-the-loop' design can consistently hold up in real environments.

Hadrian claims its solution delivers three quantifiable results: 10x improvement in critical exposure surface visibility, 80% faster remediation time, and 5x return on investment compared with manual penetration testing. These three figures are the most impactful part of the entire press release and the key to understanding the pricing logic of this funding round.

But the article gives no calculation methodology, sample size or third-party audit. The confidence level of these figures is marked at the evidence level as source claims, not confirmed facts. In other words, they describe the effects Hadrian wants the market to believe, not results that have been independently reproduced.

For a vendor claiming to be the 'only integrated' one, the lack of verifiable basis for its performance figures means outside readers cannot judge whether 10x, 80% and 5x are the lower bound, upper bound, or best-case performance in specific customer scenarios.

Hadrian's customers include U.S. and European enterprises such as McKesson, NBC Universal, Francisco Partners, TotalEnergies, Amadeus, Leroy Merlin and Damen Shipyard. These customer names appear in the press release, along with quotes from relevant executives at Leroy Merlin and Damen Shipyards, making them confirmable factual information.

Saumitra Dubey, a partner at investor SmartFin, said Hadrian's enterprise contract value growth, global customer base and retention rate are outstanding, and said it supports Hadrian in building a company with nine-figure revenue. But the specific magnitude of contract value growth, the specific retention rate figures, and the timeline for the nine-figure revenue target are all undisclosed.

Forgepoint Capital says in its introduction in the release that it manages assets of more than $1.2 billion and has an active portfolio of more than 40 companies. This information also comes from the institutional introduction in the press release and is a source claim.

Hadrian was founded by Rogier Fischer, Olivier Beg and Maurice Clin. Fischer and Beg were teenage ethical hackers, and Fischer previously founded one of Europe's earliest cryptocurrency exchanges. This background echoes the company's positioning of 'looking at security through a hacker's lens.'

The article says their goal in founding Hadrian was to build offensive security capabilities that can run continuously and at scale without stripping teams of their judgment. This statement is consistent with the aforementioned design principle of 'humans set tasks, AI executes,' but it is likewise part of the company's own narrative framework.

The post-money valuation and equity dilution of this $40M round are unclear; the article does not disclose the valuation, equity structure, specific closing timeline or the proportion of funds allocated. This means outsiders cannot judge the equity cost corresponding to the $65M in total funding, nor the specific proportion of funds going to engineering and research teams.

More critically, there is effect verification. Hadrian's claimed 10x, 80% and 5x effects lack independent verification, and their actual reproducibility is unknown; whether the statement that it is 'the only vendor integrating continuous exposure surface management and agentic penetration testing' holds cannot be judged from existing sources.

In addition, the background data cited in the article—the Anthropic disclosure, the 87% manual penetration testing figure, the 0.47% exploitable ratio—have unclear sources and statistical methods; the details and authenticity of reports about AI agents escaping test environments and intruding into corporate systems are likewise unverified in this article. Whether Hadrian can achieve the nine-figure revenue target cited by its investor, and the specific figures for its enterprise retention rate, are both undisclosed.