Rune sat down with swyx and Vibhu Sapra to discuss what’s happening as frontier AI risk become front and center. Listen to the episode on your preferred platform – link in the comments. https://lnkd.in/e_aYV5-J
Artificial Intelligence Underwriting Company (AIUC)
Technology, Information and Internet
AIUC certifies and insures AI agents.
About us
Building the trust infrastructure for AI agents. On a mission to underwrite superintelligence.
- Website
-
aiuc.com
External link for Artificial Intelligence Underwriting Company (AIUC)
- Industry
- Technology, Information and Internet
- Company size
- 11-50 employees
- Type
- Privately Held
Employees at Artificial Intelligence Underwriting Company (AIUC)
Updates
-
Artificial Intelligence Underwriting Company (AIUC) reposted this
Pandora’s box is open. AI beats humans at math - but can also escape our sandboxes. Sam, Dario, and Elon all called for frontier AI audits. The risk is no longer that models aren’t competent, but that they are all too competent. We cannot confidently deploy frontier models without oversight. Stakes are high, timelines are short. Rune and I started AIUC with a simple premise: history offers a guide for how to build confidence in new, risky technology. Electricity burned down many houses until insurers funded Underwriters Laboratories (UL) to certify lightbulbs. Today, the UL trustmark is on every lightbulb in America. The combination of standards, audits, and insurance creates an incentive flywheel that has taught the world how to trust electricity, cars and nuclear power. We must urgently build this for AI. Over the past year, a coalition has emerged fast around AIUC-1, the security standard for agents. > Frontier agent builders like Cursor, ElevenLabs, Lovable, Fin and Harvey — whose agents write production code, draft legal work, and answer customer calls for millions of people — now carry the AIUC-1 trustmark. > 250+ leaders securing institutions like JP Morgan, Fidelity, Capital One, NATO, and Palo Alto Networks, bring lessons from the trenches to our quarterly updates. > Schellman and other auditors bring traditional rigor to the audits on top of our technical evaluations. > Insurers from Lloyd’s of London bring insurance coverage to the frontier. Next, we’re extending from agents to models. The past months have shown that a strong ecosystem of standards, audits and insurance is urgently needed at the model layer. In the coming months we’re excited to share more about how to extend standards to models, quantify the risks and govern the frontier. We’re grateful for the backing of Micky and Nick from Ribbit Capital, Ali from First Harmonic, Willem from Terrain, and Nat Friedman, as we rush into the next chapter. There’s no time to waste. If you want to do your life’s work with us, we’re hiring across all roles in San Francisco.
-
AIUC is a 2026 IA40 winner. The #IA40 recognizes the 40 most important private companies in applied AI. This selection affirms a belief that we've had since we started AIUC: powerful AI needs an independent way to prove it can be trusted. Confidence, not capability, will be the binding constraint on AI adoption. That's why we're building confidence infrastructure backed by an independent standard, third-party audits, and insurance so every enterprise, bank, and government can adopt AI they can trust. Thank you to Madrona, Microsoft, NYSE, Delta Air Lines, McKinsey & Company, AWS Startups, and Google for Startups for the recognition.
-
-
Artificial Intelligence Underwriting Company (AIUC) reposted this
Spoke to The Economist about autonomous hacking. They said it best: autonomous hacking is here, governments are not ready. Today, no one is obviously liable when OpenAI's models hack Hugging Face. Liability for cyber attacks requires *intentional* hacking. OpenAI did not hack Hugging Face intentionally. Like many other places, cyber law can't make sense of autonomous agents. We're in for a confusing time as swarms of agents will scour the internet for vulnerabilities over the next 12 months.
-
-
Artificial Intelligence Underwriting Company (AIUC) reposted this
Rune and I started the Artificial Intelligence Underwriting Company to build the confidence infrastructure for AI adoption. Appreciate Fast Company AI 20 for recognizing our work. Liability rather than capability is the binding constraint on AI adoption as capabilities accelerate. Decision makers need answers to questions like: is our data secure? Is the agent robust to known jailbreaks? Can the system be misused? Fable brings these questions to the fore; this is just the beginning. We built AIUC-1, the standard for AI agents, together with 250+ Fortune 1000 security and risk leaders to build confidence in AI adoption. It measures how well agents hold up against AI agent risks including jailbreaks, hallucinations, and data leaks. And it is auditable & insurable. The future of AI is a collective effort, we’re proud to be shaping a piece of it.
-
-
Previously, businesses were forced to choose: embrace AI despite genuine risks, or wait on the sidelines while falling behind. We are changing that, so that you can move with confidence.
95% of enterprise AI pilots never make it to production. The blocker often isn’t the technology. In many situations, trust breaks down and teams have to choose between moving fast and accepting real risk, or moving slowly while competitors ship. AIUC-1 offers a third option. UiPath invited us to put their enterprise automation platform through the full evaluation: > 2,000+ tests across 10 risk categories > scenarios drawn from real incidents and current research > repeated quarterly against an evolving standard The standard covers the risks CISOs lose sleep over: data leakage, unsafe tool calls, hallucinations... and everything else that shows up in the security review. When we tested UiPath, their agents held up. They are now the first enterprise automation platform to receive AIUC-1 certification. The certification is proof that moving fast and moving responsibly don’t have to be in conflict. And that’s why we hand-delivered. Because what the AUIC-1 trust mark represents is worth handing over by a human, not as an email. Danny & I had a blast delivering it to Scott & Sheron!
-
Today, UiPath becomes the first enterprise automation agent platform to be AIUC-1 certified. Following certifications of customer service, image generation and voice agents, AIUC-1 now extends to enterprise automation agents. AIUC-1 was created to fill a critical gap - how do we verify that AI systems, especially autonomous agents, operate transparently, accountably, and safely? Traditional IT certifications weren't designed for AI's unique risks: > Prompt injection resistance > Hallucination rates under adversarial conditions > Tool call reliability > Data leakage controls > Incident response for AI-specific failures UiPath's agentic automation suite - Agents, IXP and Autopilot - held up to a rigorous audit, including 2000+ technical evaluations. This audit marks a milestone as the first AIUC-1 audit led by a third-party auditor, Schellman. See press release in the comments.
-
Previously, businesses faced a binary choice: embrace AI despite genuine risks, or wait on the sidelines while falling behind. We are changing that, so that you can move with confidence. Congratulations to the entire Intercom team!
Having an AI governance framework doesn't mean your agent is safe. That's why we run thousands of adversarial tests on the agent every quarter. Intercom went beyond telling Fin was secure - they invited us to try to break it. We run 5,000+ technical tests quarterly covering: > Real-world incidents and the latest research > The risks that block enterprise adoption: jailbreaks, tool calls, prompt injections, hallucinations Fin held up - demonstrating they’re meeting best practices. That's what earning AIUC-1 looks like - not a governance review, but proof under pressure. That is real, hard, work - and deserves more than a PDF in a shared drive. And that’s why we hand delivered.
-
When ElevenLabs earned AIUC-1 certification, we didn't just send the certificate. 5,000+ technical tests. Independent evaluation. A standard built with leading Fortune 500 enterprises deploying these agents. That's what AIUC-1 represents, and that's worth showing up for.
The pull toward AI agents is undeniable. New capabilities are emerging every day, reshaping entire industries in real time. But with that momentum comes a crisis of trust. As fast as we're hearing about breakthroughs, we're hearing about failures: leaks, rogue agents, compromised data, broken processes. For businesses of every size this creates an impossible-feeling choice: adopt and move forward with real risk, or hold back and risk being left behind. Until now, those were the only two options. AIUC-1 is changing that and here is how: ElevenLabs invited us to put their state-of-the-art agents through the most rigorous evaluation framework built: 5,000 tests across 14 risk categories. Their agent held up. And with that, they earned the first AIUC-1 certification for a voice agent. Not a badge handed out with a few clicks on a website, it is a trust mark earned through real work and competence. We delivered that credential in person, because what it represents is bigger than a PDF report or a logo on a website. It's proof that moving fast and moving responsibly don't have to be in conflict. Your business wants to move. Now it can move with confidence. Congrats Marco, Carles, Mati!
-
ElevenLabs is the first AI voice agent to achieve AIUC-1 certification. Congratulations to the entire ElevenLabs team on this important milestone! We are proud to work alongside frontier AI leaders on making agentic AI more secure, safe and reliable. Link to official blogpost in comments.
ElevenLabs just became the first voice agent to be AIUC-1 certified and insured. Enterprises want to deploy AI agents. But the question holding back adoption isn't "can it work?" - it's "can we trust it?" AIUC-1 fills that gap. A rigorous, independent certification that validates ElevenLabs' leadership in designing and deploying secure AI agents. They opened their systems to scrutiny, met the bar, and earned the right to say: our voice agents are certified and insured. Congratulations to Mati, Marco, and the entire ElevenLabs team!