Gray Swan, The AI Security Company Trusted by Every Major Frontier Lab, Raises $40M Series A

AI security company used by Anthropic, OpenAI, and Meta raises $40M to deliver on its mission to empower the world to use AI safely and securely.

PITTSBURGH, PA, UNITED STATES, May 28, 2026 /EINPresswire.com/ -- Gray Swan, the AI security company born out of Carnegie Mellon University's pioneering AI safety research team, today announced a $40 million Series A round co-led by Wing Venture Capital and Madrona with participation from Obvious Ventures, Snowflake Ventures, Hudson River Trading, Samsung Next, and existing investor Magarac Venture Partners.

Gray Swan’s mission is to empower the world to use AI safely and securely. The company has built world-leading products that help enterprises and frontier labs assess how their AI models and agents can behave in unintended ways, and then provides the security that prevents such behavior. The funding will accelerate Gray Swan's go-to-market operations, deepen its partnerships with the labs building frontier AI, and scale the team helping the world's leading organizations deploy AI quickly without compromising security.

Trusted by the labs building frontier AI.

Gray Swan is a leading security evaluator cited in 11 recent frontier model system cards, including those from Anthropic, OpenAI, and Meta. Gray Swan’s benchmarks are embedded into the safety evaluation processes for the world’s most advanced AI models before they are released to the public. No other AI security company operates at this level of access and trust across leading frontier labs simultaneously.

Every AI deployment is an open attack surface until proven otherwise.

As AI agents move from demos to production, accessing enterprise data, executing workflows, and making autonomous decisions, the attack surface has expanded rapidly. Shadow AI is proliferating inside organizations. Prompt injection, jailbreaks, and data exfiltration are now operational realities. And with global regulatory frameworks tightening, the associated costs of deploying unsecured AI are rising quickly.

Before the industry had a name for these threats, Gray Swan's founders were publishing the research that proved they were inevitable. Matt Fredrikson and Zico Kolter spent over a decade studying how AI systems fail under adversarial conditions, how those failures can be prevented, and shaping how the industry thinks about AI safety and security. As AI adoption accelerated from research tools to enterprise infrastructure, they saw the critical need to close the gap between what their research uncovered and what was available in the market to address it.

Gray Swan was built to close that gap.

“AI applications are growing at an unprecedented rate, and at Gray Swan we want to ensure that these deployments can continue without sacrificing reliability and security,” said Zico Kolter, Chief Scientist and Co-Founder of Gray Swan. “We have spent our careers conducting research into the safety of AI systems, and we are putting this into practice to ensure the companies can deploy AI with confidence.”

Gray Swan already works with over 20 customers across frontier labs and global enterprises, and has established partnerships with technology platforms, including Snowflake. As part of this partnership with Snowflake, Gray Swan’s runtime AI protection now natively integrates with Snowflake’s AI ecosystem enabling enterprises to embed Gray Swan’s AI security capabilities directly into the platform where enterprises build and scale AI applications.

"As AI agents move into production, accessing enterprise data and making autonomous decisions, runtime security isn't optional,” said Mayank Upadhyay, Chief Security and Trust Officer at Snowflake. “Gray Swan brings the same adversarial rigor trusted by the leading frontier labs directly into Snowflake, protecting customers where they're already building."

"AI is moving faster than any technology we've seen, and security hasn't kept pace," said Matt Fredrikson, CEO and Co-Founder of Gray Swan. "Our mission is to empower the world to use AI safely and securely, and this funding lets us pursue the mission on pace with the frontier. Our partnership with Snowflake is a perfect example: bringing the same security trusted by frontier labs to the platform where enterprises are already building on their data."
The Gray Swan AI Security Platform.

Gray Swan’s platform addresses AI vulnerabilities across three components: Cygnal for real-time AI protection, Shade for continuous adversarial testing, and Arena, a global AI red-teaming community that generates the threat intelligence powering both. Together, they give organizations security coverage across the full AI lifecycle.

Cygnal: Real-time AI monitoring and protection
Cygnal identifies and blocks dangerous AI model and agent behavior. Enterprises define custom policies, personalized to the specific risks their organization faces, and Cygnal enforces these policies 24/7. Cygnal reliably performs with better latency and accuracy than other solutions.

Shade: Automated adversarial testing / red-teaming
Shade is an AI red-teaming agent that automatically runs adversarial tests against AI models and agents as part of pre-deployment testing and the CI/CD pipeline. It executes attack strategies observed in the wild, surfacing vulnerabilities before they reach production.

Arena: The world’s largest AI red-teaming network.
Arena is a continuous global competition where more than 15,000 researchers and security professionals attack AI models to uncover vulnerabilities. The Arena serves as the threat intelligence backbone of the Gray Swan platform, generating more than one million high-quality, real-world attack trajectories and using them to train the models powering Cygnal and Shade; keeping defenses current with the bleeding edge of AI research.

"Only a handful of security vendors count frontier labs as customers. Gray Swan is even more than that — named in their system cards, embedded in the safety process before the world's most advanced models go live. Enterprise security teams are now applying that same bar to every AI deployment they run. Gray Swan already sets it," said Jake Flomenberg, Partner at Wing Venture Capital.

"I've spent the past few years working alongside the tech leaders who actually have to make AI security decisions — CISOs in our advisory network, hyperscalers, and the frontier labs themselves." Said Vivek Ramaswami, Partner at Madrona. "I have never heard such a chorus of praise for one company. Gray Swan is the name every group gives you, and the practitioners using the product describe it the same way: years ahead of anything else they've evaluated. Matt, Zico, and the team did the research that defined AI security, and they're now building it into the practical reality the field needs."

Samantha Link
Gray Swan
press@grayswan.ai
Visit us on social media:
LinkedIn
X

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.