Businesses across industries and geographies are racing to develop, deploy, and use AI. Winning responsibly means embedding compliance early, before regulators and clients demand it. Awareness of AI risk is high: 4 out of 5 companies that use AI are getting questions from customers about risk, and for good reason. Data breaches are expensive, but the ultimate cost of misunderstanding AI’s impact is reputation. Once customer trust is broken, it’s next to impossible to mend. Even so, most organizations don’t know where to start. 

According to A-LIGN’s 2026 Compliance Benchmark Report, concern is growing: 72% of companies worry about AI’s impact on compliance, up from 58% in 2025. Agentic AI raises the stakes further as these systems don’t just answer prompts. They take automatic, multi-step actions inside real environments. Traditional frameworks weren’t built for that kind of automation. Agents need a standard designed for agentic safety and security, one that proves controls hold when systems act on their own. 

Ungoverned agents create unseen risk 

Businesses deploy agents faster than they can secure them. Shadow AI, misconfigured agents, and hallucinations create risk that teams can’t see and can’t forecast. Leaders need tangible, data-driven benchmarks, not guesswork. 

No standard means stalled deals 

Without a shared standard, trust runs through lengthy, bespoke questionnaires. Sales cycles stretch. Revenue waits. Organizations need an industry benchmark that works as a credibility marker. 

AIUC-1 

AIUC-1 is the world’s first AI agentic security, safety, and reliability standard. Built with 100+ Fortune 500 CISOs, it covers data and privacy, security, safety, reliability, accountability, and societal risk. It operationalizes emerging AI legislation and best practices. It updates quarterly, so the standard keeps pace with technology, risk, and regulation. 

AIUC-1 is built for organizations that are: 

• Adopting or developing agentic AI systems 
• Applying agentic AI in high-risk, high-visibility use cases 
• Proactively demonstrating trust and secure AI practices 

Common in-scope AI agent systems include customer-facing support agents, sales and marketing AI agents, internal workflow automation agents, DevOps and IT automation, voice and multimodal agents, and AI systems with access to sensitive data or actions. 

How AIUC-1 tests your agents 

1. Context first: AIUC-1 studies the environment where your agents operate. 
2. Tailored evaluations: From its proprietary taxonomy of AI risks and attacks, it selects 2,000 to 4,000 evaluations that mirror your agent’s exposure. 
3. Real pressure: Evaluations range from normal user behavior to sophisticated attacks. AIUC-1 simulates them, collects the agent’s output, and grades each result pass or fail. 
4. Severity that matters: Results fall into five severity levels. P0 and P1 findings require immediate remediation, and AIUC-1 can’t issue certification until you resolve them. P3 and P4 findings are expected. No company is free of them. 

Testing happens before certification and every quarter after. Requirements span technical safeguards, operational processes, and legal and governance controls, and each is mandatory or optional based on agent scope. 

Continuous vigilance, not a point-in-time snapshot 

AIUC-1 is an annual certification with ongoing quarterly testing. Certified organizations gain continuous visibility into their agentic environment. They spot risk early, remediate fast, and show up-to-date compliance. 

That visibility pays off in multiple ways, it gives leadership and boards internal assurance that agent risk is under control, and it gives customers and partners a trust signal they can rely on. AIUC-1 is the new benchmark for agentic AI security. Certified organizations move first, shorten procurement, and earn credibility in the market. 

Who’s behind AIUC-1 

AI Underwriter Company (AIUC) was founded by veterans of Anthropic, McKinsey, and AI safety research. Its consortium includes 100+ technical contributors from security organizations, private corporations, government, and the world’s largest consultancies. AIUC also raised a $15M seed round led by Nat Friedman, former GitHub CEO, and an additional $40M from Ribbit and First Harmonic. 

Why A-LIGN 

A-LIGN brings proven rigor to AIUC-1: 5,900+ ISO audits, 13,800+ SOC 2 assessments, and a 100% FedRAMP authorization success rate. We tailor every engagement to your industry, from financial services and tech to healthcare and federal. Our auditors go beyond checkbox compliance to deliver strategic engagements that support audit consolidation, tech-enabled efficiency, and scale.

The A-LIGN difference for AIUC-1 brings:

• Unmatched scale and experience: A-LIGN is the largest SOC 2 auditor in the world, with 20+ years of experience across frameworks such as ISO 27001, ISO 42001, FedRAMP, PCI and more.
• AI backed by the expertise of 36k+ audits: A‑SCEND, our audit management platform, is designed to help teams save meaningful time, especially during audit preparation, without compromising rigor. The platform seamlessly combines technology and human expertise across the entire audit lifecycle.
• AI governance expertise: Our experts have hands-on experience delivering multiple AI governance services, so can tailor the right solution for you.
• A quality-first approach: Our methodology and approach is designed to be collaborative engagements that go beyond checkbox compliance.

Contact us today to get started.