Artificial Intelligence Safety Startup Raises $40 Million in Funding

Serdar HocamAuthor & Editor

Founded by a former Anthropic employee and a former METR executive, the Artificial Intelligence Underwriting Company has raised funding to provide an auditing layer for enterprise AIs.

◉ 0 views
Early Anthropic hire, former METR COO have found a way to rein in rogue AI agents | TechCrunch

Founded by alumni of Anthropic and METR, the Artificial Intelligence Underwriting Company has secured $40 million in a round led by Ribbit Capital to audit and certify enterprise AI agents.

Founding and Founders

Rune Kvist, an early employee at Anthropic, and Rajiv Dattani, the former chief operating officer of METR, have launched a new startup called the Artificial Intelligence Underwriting Company.

The two co-founders, who are relatives, stated that they aim to provide security solutions for companies developing AI models and institutions.

Details of the Investment Round

A $40 million Series A investment round led by Ribbit Capital, with participation from First Harmonic, has been completed.

Having previously raised $15 million in its seed stage, the startup has reached a total of $55 million in financing with this round.

Developed Audit and Certification Layer

The new startup has built a third-party auditing and certification layer by adopting a cybersecurity model for AI risks.

Developed with reference to the SOC 2 standard, the AIUC-1 standard and testing service enable companies to independently evaluate their AI agents.

Testing Processes and Reporting

Shaped by a consortium of approximately 250 security and risk leaders, the tests include five thousand evaluations covering scenarios such as jailbreaking, data leakage, and hallucinations.

As a result of the tests, one-hundred-page reports detailing the reliability of the agents are prepared, while the process is carried out through human oversight and AI analysis.