AI startup Gray Swan has secured $40 million in Series A funding, valuing the company at $200 million. The startup employs approximately 15,000 security professionals who act as hackers to pressure-test AI models from major labs like OpenAI and Anthropic. This data is used to train Gray Swan's AI agents, Shade and Cygnal, which identify and mitigate vulnerabilities in AI systems, preventing the generation of harmful content or unauthorized tool access. The company's founders, professors Matt Fredrikson and Zico Kolter, have a history of AI safety research, including discovering a significant jailbreaking technique. AI
IMPACT This funding and its focus on AI safety testing could accelerate the development and deployment of more secure AI models across the industry.
RANK_REASON Significant funding round for an AI safety company with major lab customers. [lever_c_demoted from significant: ic=1 ai=0.7]
- Anthropic
- ByteDance
- Claude Mythos
- Gemini
- Google Deepmind
- GPT-5
- Gray Swan
- Hudson River Trading
- Kameron Bettridge
- Madrona
- Matt Fredrikson
- Meta
- OpenAI
- Samsung Next
- Snowflake Ventures
- Wing VC
- xAI
- Zico Kolter
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
