TechCrunch reports Artificial Intelligence Underwriting Company launched AIUC-1, a third-party audit and testing service for agents modelled on SOC 2. Ribbit led a 40 million dollar raise the paper calls Series A. Founders Rune Kvist and Rajiv Dattani name Harvey, ElevenLabs, Cursor and Lovable among buyers. The Verge and Wired pages and feeds checked on 15 September did not carry the same round.
Artificial Intelligence··Night
A SOC 2 pattern for agents, with a Series A
TechCrunch on 15 September 2026 reports Artificial Intelligence Underwriting Company launched AIUC-1 as a third-party audit and testing service. The firm says Ribbit led a raise that brought in 40 million dollars. Rune Kvist came from early Anthropic; Rajiv Dattani had been METR’s operations chief. In the interview, Kvist argued the stall is not model IQ: lenders, clinics, ministries and armed forces withhold agents when no underwriter will back the pledges those buyers already sold downstream. The standard copies the SOC 2 cybersecurity pattern.[1]
250 buyers, named customers, 55 million dollars total
TechCrunch says about 250 security and risk leaders who buy agents helped shape the standard. The paper lists Harvey, ElevenLabs, Cursor and Lovable among buyers. A 15 million dollar seed sat underneath, with Nat Friedman’s NFDG, Emergence and Terrain in that earlier round, plus Anthropic co-founder Ben Mann. Combined capital reaches 55 million dollars. Those sums are company and investor statements, not an independent audit result.[1]
One report for the audit layer and the round
AIUC-1 wants written agent commitments on a SOC 2 pattern; the 40 million dollar Ribbit raise is in the same TechCrunch report. The 250 buyers and 55 million dollars total come from that text. No second publisher confirmed this round on its own page on 15 September.[1]
