HiddenLayer Raises $100 Million as AI System Security Accelerates

Just three years ago, when the AI security startup HiddenLayer raised its first funding round of $50 million, the biggest question in the field was whether the threats the company was addressing would materialize at a scale large enough to make it a real market. At the time, it was hard to find concrete examples of attacks targeting AI systems at scale. But the landscape shifted with remarkable speed, and securing these systems has become an urgent necessity that companies are racing to address.

An Exploding Market

Today, security companies are racing to build products capable of monitoring not just intelligent agents, but also the tools and plugins they use. And while there have been few headlines about these agents actually being exploited, the risk of them going rogue during real-world operation remains real and present. This has been clearly reflected in spending levels, with Gartner estimating that companies will spend around $2.83 billion this year on products to secure AI tools—an 83% increase over the previous year—with projections that the figure will jump to nearly $4.78 billion next year.

Exceptional Revenue Growth

HiddenLayer develops tools to protect AI models, agents, and workflows from adversarial attacks, vulnerabilities, and malicious code injection, and the company has benefited greatly from this market shift. According to co-founder and CEO Chris Sestito, the company's annual recurring revenue has grown more than tenfold over the past year, reaching "tens of millions" of dollars, with more than 90% of this growth coming from new customers who signed their contracts within the same year.

Financial services companies and major tech firms building AI products are currently the company's most prominent sectors, in addition to contracts with the U.S. Department of Defense and the intelligence community. One of its customers is said to be a leading provider of advanced models with "more than 700 million weekly users."

Details of the Funding Round

To capitalize on this momentum, the Austin-based startup raised $100 million in a Series B round led by Delta-v Capital, with participation from a number of parties, most notably:

• Ten Eleven Ventures
• Morgan Stanley
• M12, Microsoft's investment arm
• Booz Allen Hamilton

Expanding Scope Rather Than Pivoting

Sestito emphasizes that the company is still selling the core of what it offered in 2023, but the biggest change has been expanding its existing products—from detection, runtime protection, attack simulation, and supply chain security—to include defending against prompt injection, agent manipulation, and malicious tool use. He explains: "Inference remains inference; whether it's on a traditional machine learning model, generative AI, or an agent-based workflow, we didn't have to change our course, but rather expand our scope."

He stressed that runtime protection has become a top priority as AI systems spread throughout business, likening it to endpoint detection and response solutions, but tailored specifically for AI.

The Risk of Open-Source Models

Sestito highlighted a new opportunity that attackers exploit lying in open-source models, explaining that the company scans about 50 different frameworks for AI files to ensure that the tool a user is dealing with is actually what it claims to be, and to detect cases where models are disguised inside other models.

Expansion Ambitions and Challenges

The company will direct the new funding toward strengthening sales and distribution, while continuing to grow its engineering and research teams, in addition to expanding into Europe and the Middle East and Africa region. However, HiddenLayer faces strong competition, as major cybersecurity companies such as Cisco and Palo Alto Networks prefer to buy rather than build, while competing startups have each raised more than $100 million in funding.

Sestito believes that his company's mission now is to "expand vertically in parallel with the growth of AI," then later expand horizontally toward areas of cybersecurity that increasingly rely on AI. It's an ambitious goal, but the company must first prove its ability to convert its head start into a sustainable business before the rest of the players in this crowded field catch up.

✦ بقلم فريق دروب أيديا