Editorial illustration for AIUC Raises USD 40 Million to Develop Rogue AI Controls
AIUC Raises $40M for AI Safety Controls
AIUC Raises USD 40 Million to Develop Rogue AI Controls
Rune Kvist and Rajiv Dattani are brothers-in-law with a shared résumé in AI safety, and on Tuesday they announced $40 million to bet that enterprises will pay to keep AI agents in line. The Series A, led by Ribbit Capital with First Harmonic joining in, pushes their startup's total funding to $55 million, on top of a $15 million seed round backed by Nat Friedman's NFDG, Emergence, Terrain, and Anthropic co-founder Ben Mann.
Kvist was an early employee at Anthropic. Dattani ran operations as COO of METR, the AI safety research group. Their company, Artificial Intelligence Underwriting Company, or AIUC, already counts Cursor, Lovable, Harvey, and ElevenLabs as customers.
The pitch: as AI models get smarter, they don't get easier to manage. They get harder.
That premise landed the day after Anthropic researcher Jacob Coxon resigned, citing fears that AI could pose an existential threat by decade's end. AIUC's answer isn't a kill switch. It's an audit and certification system borrowed from an older industry playbook, one built for institutions that can't afford to guess.
Using the widely adopted cybersecurity standard SOC 2 as its muse, AIUC has developed a standard called AIUC-1 and a testing service to validate agents against the standard.
Why this matters
Kvist and Dattani are betting that enterprises will pay for AI insurance before regulators force the issue, and $55 million total from Ribbit, First Harmonic, and Nat Friedman's NFDG suggests investors think that bet is sound. But there's a tension worth sitting with: the same week AIUC closes its Series A, a researcher at Anthropic, one of the labs whose models AIUC presumably wants to certify as safe, walked away citing existential risk. That's not a great advertisement for "we've figured out the controls layer."
For founders building on top of frontier models, AIUC's pitch is straightforward: agents fail in enterprise environments in ways that look like security incidents, not philosophy problems, and someone needs to sell the fix. That's a real market. Whether it's the same market as "rogue AI," or whether AIUC is just doing enterprise risk management with better branding, is the question we'd push them on. Worth watching whether their customers are buying peace of mind or an actual technical guarantee, because those are very different products.
Related Reading
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
From Archives Musk antitrust suit against OpenAI proceeds after Apple resolves ChatGPT dispute
From Archives Anthropic Researcher Quits, Questions AI Builders' Beliefs
Common Questions Answered
What is AIUC-1 and how does it relate to AI agent safety?
AIUC-1 is a safety standard developed by AIUC that was modeled after the widely adopted cybersecurity standard SOC 2. The company has created both the AIUC-1 standard and a testing service that allows enterprises to validate their AI agents against this standard to ensure they remain under control and operate safely.
Who are the founders of AIUC and what is their background in AI safety?
AIUC was founded by brothers-in-law Rune Kvist and Rajiv Dattani, who both have extensive experience in AI safety. Kvist was an early employee at Anthropic, while Dattani previously served as COO at METR, bringing operational and safety expertise to the venture.
How much funding has AIUC raised and who are the key investors?
AIUC has raised a total of $55 million across two funding rounds. The Series A round of $40 million was led by Ribbit Capital with First Harmonic joining in, while the earlier $15 million seed round was backed by Nat Friedman's NFDG, Emergence, Terrain, and Anthropic co-founder Ben Mann.
What is the business model behind AIUC's approach to AI safety?
AIUC is betting that enterprises will proactively pay for AI safety controls and certification before regulators mandate such requirements. The company plans to offer testing and validation services that help organizations ensure their AI agents comply with the AIUC-1 standard, positioning itself as an AI safety insurance provider.
Why is there tension around AIUC's mission despite strong investor backing?
While AIUC's Series A closing attracted significant investment from major backers, the same week an Anthropic researcher departed citing existential risk concerns about AI. This creates a contradiction where AIUC aims to certify models as safe while one of its partner labs' own researchers expresses fundamental concerns about AI safety.
Further Reading
- Early Anthropic hire, former METR COO have found a way to rein in rogue AI agents - TechCrunch
- Artificial Intelligence Underwriting Company raises $40M to insure AI agents - CryptoBriefing
- AIUC, a startup creating insurance for AI agents, emerges from stealth with $15 million seed - Fortune
- Enterprise AI Agent Funding Surges to $435M in Five Months - Yahoo Finance
- HiddenLayer Raises $100 Million for AI Runtime Security - SecurityWeek
