Life’s good out on the frontier 🤠
Incredible work by the research team on training a best-in-class cybersecurity model. So excited for what’s next
Today we're announcing dfs-large1, our newest cybersecurity model that achieves best-in-class performance on vulnerability detection tasks.
Besides frontier AI labs, only a handful of companies have built specialized models that reach the state of the art in their domain. We're proud to be the first to do it for cybersecurity.
dfs-large1 is built on GLM-5.2 and post-trained with reinforcement learning inside depthfirst's security infrastructure. We evaluated it on depthfirst-bench, our benchmark of long-horizon vulnerability discovery across complex repositories, where it achieves best-in-class performance.
Training improvements have not plateaued yet and we expect additional performance gains as we continue training.
A huge thank you to @FireworksAI_HQ for being an outstanding training partner. Their infrastructure enabled us run large-scale reinforcement learning efficiently and iterate much faster.
dfs-large1 is now in preview within the @depthfirstlabs platform
1
12
1,779
steve retweeted
Today, we’re introducing issue ingestion in depthfirst. Security teams can bring findings from scanners, bug bounties, or external audits. depthfirst agents check each finding against your current code and explain whether it’s exploitable. They also merge duplicate reports into one issue and rate severity using your organization’s threat model.
Issue ingestion is now in public preview via the depthfirst UI, REST API, and MCP.
Read the blog: depthfirst.com/post/issue-in…
steve retweeted
Today, we’re disclosing phpAfterShell (CVE-2026-126714), a critical vulnerability in PHP discovered by depthfirst researchers using the depthfirst platform using a combination of models including our in house model.
The vulnerability affects how PHP handles file-based sessions. Under default configurations, an unauthenticated attacker could exploit it to execute code on the server. It has a CVSS v4.0 score of 9.2.
We reported the issue privately to the PHP team and worked with the maintainers on developing the patch. We recommend upgrading to the latest patched version of PHP as soon as possible or turn off session upload progress as a temporary mitigation if upgrading requires more time.
To help organizations assess their exposure, we’re offering language runtime scans of repositories and container image scans to identify potentially vulnerable PHP versions. Our team can also help assess temporary mitigations for organizations that can’t upgrade immediately.
Read more about phpAfterShell and request a scan: phpaftershell.com
Technical writeup: depthfirst.com/research/phpa…
steve retweeted
CVSS 9.9 in Oracle Database. Oracle just patched CVE-2026-114819, which I reported. Chained with a GraphQL bug I disclosed in October, it turns an exposed HTTP endpoint into command execution on the database host.
Then I wondered if Oracle was the exception. It wasn't. PostgreSQL and MySQL both fell to full RCE from the lowest-privilege accounts you can create.
These were meant for Pwn2Own, but we failed to register. So you get the write-up instead.
depthfirst.com/research/narr…
Your “low severity” findings may only be that in isolation
Vulnerability chaining is now available in depthfirst.
What do you miss when you review vulnerabilities one at a time?
An issue that looks low-risk in isolation can become part of a critical exploit when chained with others.
In our recent work on V8, Chrome’s JavaScript engine, we found a bug that let us modify memory inside V8’s sandbox. The sandbox is designed to contain that kind of damage. But we also found a second bug that, combined with the first, let us escape the V8 sandbox and run native code in Chrome’s renderer process.
Our latest post explains how we identify and validate these chains so teams can see when a finding that looks minor on its own is part of a more serious attack path.
depthfirst.com/post/vulnerab…
3
94
steve retweeted
Coding agents work best when they can run the full application and verify their own work. For complex applications, this means giving agents access to databases, background workflows, Kubernetes clusters, and everything else they need to test their changes.
At depthfirst, we built our own secure cloud development environment in less than a week using AWS, Tailscale, Nix, and depthfirst.
Each developer gets a persistent devbox, and each task runs in its own isolated environment, allowing agents to work in parallel. Everything stays within our private AWS network, with access controlled through Tailscale and package requests checked by depthfirst's Dependency Firewall in configured build and registry paths. Engineers can switch between Codex, Claude Code, and other agents without losing context.
@andreamichi and @seshank_k share how we built it and how other engineering teams can do the same.
depthfirst.com/post/secure-a…
A brief glimpse into the future of the development lifecycle
Every company should own their secure cloud dev environment where agents can change code, run the full application, and verify their work. At @depthfirstlabs we care tremendously about this and today we share how we built our environment.
Each developer gets a persistent devbox, and each task gets its own application stack, database, and workflow state, so agents work in parallel. Engineers have flexibility to switch between Codex, Claude Code, and other agents without losing context.
The environment runs inside our private AWS network, with access controlled through @Tailscale and package requests checked by depthfirst’s Dependency Firewall in configured build and registry paths.
When an agent finishes an iteration, it can share reproduction steps, video evidence, and a private preview URL. We can open the exact environment it tested and review the behavior ourselves.
We had it operational within a week.
5
266
steve retweeted
Running a lower-cost model multiple times can match the vulnerability detection rate of frontier models at a small fraction of the cost.
On a 30-task subset of dfbench, five runs of DeepSeek v4.1 Flash found about 71% of known vulnerabilities, up from about 52% with a single run, at $4.24 per task.
Running it 15 times increased recall further to 80%, roughly 13% of Mythos 5’s cost, which found approximately 69% of the vulnerabilities.
Because LLMs are stochastic, they can find different vulnerabilities each time they analyze the same codebase. Combining findings across runs increases coverage, but can also introduce more false positives.
To control that noise, we used our harness to review and deduplicate findings across runs, then filter them by confidence. In separate validation experiments, this approach generally improves both recall and precision.
These results suggest that repeated sampling can improve the tradeoff between coverage, precision, and cost. The gains depend on the model, the number of runs, and how findings are filtered.
Read the research: depthfirst.com/research/bon-…
Seems clear in hindsight but at the time it wasn’t as obvious. These predictions were dead on though + it’s awesome to be at the forefront of the revolution happening in cybersecurity
When @andreamichi, Daniele, and I started planning on founding @depthfirstlabs, we wrote a document outlining our vision for the future of cybersecurity and our company’s strategy. Today, 2.5 years later, cybersecurity has become one of the most important areas of AI development and beyond.
I recently reread the document and thought it would be interesting to share how our predictions held up and how we have executed on our plan. I’m sharing the text from the original Google Doc we wrote, last modified on June 18th, 2024.
depthfirst.com/post/original…
Our original thesis was that AI would fundamentally change the economics of vulnerability discovery and exploitation. Finding sophisticated vulnerabilities had historically required scarce expertise and considerable time. We believed AI would dramatically reduce those constraints, giving attackers the ability to find and exploit vulnerabilities at a scale that existing defenses would struggle to keep up with. Today, increasingly capable AI models are discovering vulnerabilities that have gone undetected for years, and powerful open-weight models are making these capabilities accessible to virtually anyone.
One passage from the document feels particularly prescient today: “In a world of abundance intelligence, this situation will not be tenable and will lead to catastrophic consequences if not addressed. The company’s mission is to identify and patch all the vulnerabilities in the software running the world.”
We also believed that AI would make it possible (and necessary) to build defensive systems capable of finding and fixing vulnerabilities at a scale that had never been economically feasible, and set the objective of building “computer-security artificial general intelligence”.
Two years in, we have raised $120m, built a growing team of 70, trained a SOTA cybersecurity model, found vulnerabilities in the biggest open-source projects and major enterprises, and have an eight-figure revenue run rate. AI startups, Fortune 500 companies, and government agencies use depthfirst to reduce risk across their systems and meet today’s security challenges.
We’re growing our team and are looking for excellent, mission driven people across engineering, GTM, and ops. If you’re excited about our mission, consider applying.
6
305
steve retweeted
We are showcasing our work with @nvidia to post-train Nemotron 3.5 Lightning for defensive cybersecurity at GTC.
As AI makes cyberattacks faster, cheaper, and more common, we need to equip defenders with advanced cyber intelligence that is affordable enough to be always on.
That's what we're building at depthfirst: powerful, specialized security models that defenders can run on their own infrastructure, with full control over their models and data, at a fraction of today's cost.
We're proud to work with NVIDIA to strengthen the U.S. AI cybersecurity ecosystem and help put sovereign security intelligence in the hands of defenders.
steve retweeted
As part of our Open Defense Initiative, we used dfs-large1 from @depthfirstlabs to uncover two memory corruption bugs and a sandbox escape in Chrome’s V8 JavaScript engine. One bug had been present for more than eight years.
We combined each memory corruption bug with the sandbox escape to build two separate exploit chains and won 2 v8ctf slots!
Ram doing excellent benchmarking work, uncovering the full cyber potential of the latest frontier models
We tested Gemini 4 Argon on @depthfirstlabs dfbench, our benchmark for finding real vulnerabilities in code.
On application and smart-contract tasks, where it rarely refused, it found 74% of known vulnerabilities, the highest recall we have measured. Across all 165 tasks it scored 67.5%, second only to Mythos 5 (69%{.
6
205




























