Organisation · company · US

Haize Labs

Automated red-teaming and large-scale adversarial stress-testing of AI models.

Reference profile

Haize Labs is a US startup that automates the stress-testing of AI systems, generating adversarial inputs at scale to surface the failures and jailbreaks a model will produce before it is deployed. Founded in 2023 by Leonard Tang, Richard Liu and Steve Li — who met as Harvard undergraduates — it markets its "haizing" as a way for labs and enterprises to run structured red teaming rather than relying on ad-hoc manual testing. The company raised a $12.5m seed round in 2024 led by General Catalyst, reported at a $100m valuation, and has said it holds contracts with Anthropic, Scale AI and AI21.

Category
Safety & alignment research
Founded
2023
HQ
New York, US
Funding
$12.5m seed (2024), led by General Catalyst, reported at a ~$100m valuation