Anthropic and former METR COO launched AIUC, a startup focused on AI safety for enterprises. The company aims to prevent rogue AI agents by applying cybersecurity standards to AI risk management.
AIUC reported $40 million in Series A funding led by Ribbit Capital, bringing its total funding to $55 million. The round included participation from First Harmonic and previous seed investors like Nat Friedman and Ben Mann.
AIUC built a third-party audit and certification layer for AI agents, using a standard called AIUC-1. The company developed a testing service to validate agents against the standard, with tests covering jailbreaks, hallucinations, and data leaks.
"Banks, hospitals, governments and militaries no longer decline to deploy AI because a model isn’t smart enough," said Rune Kvist, an early Anthropic employee. "They decline because they’ve made commitments to their own customers about what a system will and won’t do, and nobody can currently guarantee that."
The announcement follows Anthropic researcher Jacob Coxon’s resignation over concerns about AI risks. AIUC’s approach mirrors METR’s work, which OpenAI previously used to investigate its Hugging Face incident.
AIUC did not say how it will scale its testing framework, and it remains unclear how widely the AIUC-1 standard will be adopted. The company plans to release its first audit reports in the coming months.
Source: techcrunch
