depthfirst, an applied AI lab focused on software security, has raised $80 million in a Series B round. The round was led by Meritech Capital, with participation from Forerunner Ventures and The House Fund, alongside existing backers including Accel, Box Group, Liquid 2 Ventures, Alt Capital, and Mantis VC. The new capital brings the company’s total funding to $120 million.

What makes the timing notable is the pace. The Series B comes less than 90 days after depthfirst emerged from stealth with $40 million in Series A funding. That kind of back-to-back raise signals strong investor confidence, not just in the company’s product, but in the problem it is trying to solve.

A Security Model Built In-House

Alongside the funding news, depthfirst announced dfs-mini1, its first proprietary AI security model. The model focuses initially on securing cryptocurrency smart contracts. It was built on an open source foundation, then post-trained through reinforcement learning in security-specific environments. The company evaluated it against OpenAI EVMBench, a standard benchmark for detecting smart contract vulnerabilities.

The early results are hard to ignore. In initial testing, dfs-mini1 outperformed frontier models while running at 10 to 30 times lower cost. Internal evaluations also suggest the model generalizes beyond smart contracts, showing early signs that the training approach carries over to other areas of security.

“When you own the training process, you can optimize for what actually matters in your domain. In our case, that means vulnerability detection and verification. The result is a model that can be cheaper to run, better at the task, more responsive to continued investment than a general-purpose system.”

Andrea Michi, Chief Technology Officer of depthfirst

The Problem Legacy Security Cannot Keep Up With

AI has changed how software is built. Companies across industries are shipping more code, more services, and more infrastructure updates than ever before. The scale is hard to match with traditional security tools. At the same time, attackers are using AI to find weaknesses faster and run operations at a scale that older defense systems were never designed to handle.

depthfirst was built to address exactly this gap. Its AI-native platform reasons across entire software systems, identifies real risk, and delivers specific fixes directly inside developer workflows. The goal is to help teams secure software at the same speed they build it, rather than treating security as a separate process that slows everything down.

“Recent public-market reactions suggest investors are starting to recognize that AI will disrupt the legacy security stack. But to win in security, companies will need to deploy security-specific models in products optimized for real security workflows. To build these models, you need specialized data, domain-specific evaluation, and deep expertise in post-training. Our team is one of the few in security able to do that.”

Qasim Mithani, co-founder and CEO of depthfirst

Where the Funding Goes

The $80 million will go toward training additional security models across new domains, growing the AI research team, and scaling enterprise adoption. depthfirst is building out both sides of the equation: the product layer that customers use, and the intelligence layer underneath it.

Since making its platform generally available in late 2025, the company has partnered with Fortune 500 companies and fast-growing software businesses including ClickUp, Lovable, Supabase, incident.io, and Moveworks. The platform already shows strong performance in production — 80 percent of its fix recommendations are accepted and merged by developers, and the company reports significantly higher recall with far less noise compared to competing approaches.

About depthfirst

depthfirst was founded on a specific thesis: that winning in high-stakes AI domains requires two things. First, products that solve real customer problems. Second, specialized models built for the most critical tasks within each domain. The company’s platform, called General Security Intelligence, builds context on a company’s code, infrastructure, and business logic. It finds complex vulnerabilities, filters out the noise, and gives developers fixes that are ready to merge.

The company’s mission is to help secure the software that powers critical systems before AI-empowered attackers get there first.

Investors Back the Combination of Research and Security Experience

Meritech Capital led the round. General Partner Arsham Memarzadeh pointed to the team’s rare combination of research capability and security domain knowledge as a key factor. “Securing modern software systems is more important than ever with the increased sophistication of AI-enabled attacks,” Memarzadeh said. “What impressed us about depthfirst is their rare combination of research capabilities and security experience that has helped them deliver the exact solution customers need.”

Sara Ittelson, Partner at Accel, emphasized the company’s approach to owning both the product and the intelligence layer. “What stands out about depthfirst is that they are not stopping at the application layer,” Ittelson said. “They are building a company that can own both the product experience and the intelligence underneath it. We continue to believe this approach will fundamentally change how modern systems are secured.”

Other investors in the round include Forerunner Ventures, The House Fund, Box Group, Liquid 2 Ventures, Alt Capital, and Mantis VC.

Origianl Creator: Ekaterina Pisareva
Original Link: https://justainews.com/companies/funding-news/depthfirst-closes-80m-to-expand-ai-security-platform/
Originally Posted: Tue, 31 Mar 2026 21:08:59 +0000