An early Anthropic employee and the former Chief Operating Officer of the AI safety research organization METR have launched a startup aimed at reining in rogue AI agents within enterprises. Rune Kvist and Rajiv Dattani established the Artificial Intelligence Underwriting Company (AIUC), securing a combined $55 million in funding, including a $40 million Series A round.
The startup addresses a critical barrier in enterprise technology adoption: the inability to guarantee predictable behavior from complex autonomous AI systems.
By SarmayaNext AI & Emerging Tech Desk • ✓ Fact-Checked • Published September 2026
What Are the Developments Surrounding the Early Anthropic Hire and Former METR COO?
An early Anthropic hire and former METR COO founded the Artificial Intelligence Underwriting Company (AIUC) to provide third-party audits and certifications for enterprise AI agents using a rigorous 5,000-test evaluation framework.
The launch of AIUC follows growing industry scrutiny over autonomous artificial intelligence systems. Rune Kvist, an early employee at Anthropic, partnered with Rajiv Dattani, former COO of AI safety research group METR, to tackle the challenge of making advanced AI models safer and easier to control inside commercial environments.
On Tuesday, AIUC announced the close of a $40 million Series A funding round led by Ribbit Capital, with participation from First Harmonic. This follows an earlier $15 million seed round backed by Nat Friedman’s fund NFDG, Emergence, Terrain, and Anthropic co-founder Ben Mann, lifting the startup’s total capital raised to $55 million. The company currently counts prominent AI firms such as Cursor, Lovable, Harvey, and ElevenLabs among its customer base.
AIUC Funding Overview
| Funding Round | Amount | Key Investors |
|---|---|---|
| Seed Round | $15 Million | Nat Friedman (NFDG), Emergence, Terrain, Ben Mann |
| Series A | $40 Million | Ribbit Capital, First Harmonic |
| Total Funding | $55 Million | Multiple institutional and angel investors |
AIUC Agent Audit Workflow
- Consultation with a 250-member consortium of enterprise security and risk leaders to establish testing criteria.
- Execution of approximately 5,000 automated tests assessing jailbreaks, hallucinations, and data leaks.
- AI-driven analysis of test behavior paired with mandatory human verification.
- Issuance of a comprehensive ~100-page performance and reliability audit report.
How Enterprise AI Auditing Shapes Governance and Adoption
The core thesis behind AIUC is that organizations—ranging from commercial banks and hospitals to government agencies and military branches—often restrict AI deployment not due to a lack of intelligence in the underlying models, but because they cannot guarantee adherence to customer commitments and safety boundaries.
Drawing inspiration from the cybersecurity standard SOC 2, AIUC has introduced a specialized evaluation framework named AIUC-1. To formulate this standard, the startup convened a consortium of approximately 250 enterprise security and risk leaders who purchase and deploy AI agents. By soliciting monthly feedback from these buyers, AIUC designs testing protocols that evaluate models against jailbreaks, hallucinations, and data leakage scenarios across roughly 5,000 individual test cases. The auditing process relies on automated AI agents to execute tests and analyze data, concluded by human verification to generate a comprehensive 100-page performance and reliability report.
Key Takeaways
- Founders Rune Kvist and Rajiv Dattani bring deep experience from Anthropic and METR to enterprise AI governance.
- AIUC secured $55 million in total funding, highlighted by a $40 million Series A led by Ribbit Capital.
- The startup developed the AIUC-1 standard and a 5,000-test audit process modeled after cybersecurity frameworks like SOC 2.
- Enterprise customers include notable AI platforms such as Cursor, Lovable, Harvey, and ElevenLabs.
The Insider Take
The transition of safety experts from research labs to commercial auditing reflects a broader market maturation where enterprise buyers demand verifiable compliance over raw model capability. By establishing third-party certification layers comparable to established IT security frameworks, startups like AIUC are bridging the trust gap necessary for institutional and regulatory acceptance of autonomous software agents.
Frequently Asked Questions About Early Anthropic hire former
What is the background of the founders behind AIUC?
AIUC was founded by Rune Kvist, an early employee at Anthropic, and Rajiv Dattani, the former Chief Operating Officer of the AI safety research organization METR.
How much funding has AIUC raised to date?
AIUC has raised a total of $55 million, comprising a $15 million seed round and a $40 million Series A round led by Ribbit Capital.
What auditing standard does AIUC use for evaluating AI agents?
AIUC uses a proprietary standard called AIUC-1, which is modeled after the cybersecurity standard SOC 2 and evaluates agents through a suite of roughly 5,000 tests.
PS: For educational and informational purposes only. Technology specifications and availability are subject to regional rollout and device compatibility.