AIUC-1 is a standard for AI agent security, safety and reliability, created by the Artificial Intelligence Underwriting Company (AIUC) and developed with the AIUC-1 Consortium. Its requirements cover six areas: data and privacy, security, safety, reliability, accountability and society.

How often is AIUC-1 updated?

Every quarter, on January 15, April 15, July 15 and October 15. AIUC documents each change in a public changelog.

How is AIUC-1 different from SOC 2 and ISO 42001?

AIUC says SOC 2 is not AI-specific and does not cover risks unique to AI, and that ISO 42001 focuses on AI governance and management systems while AIUC-1 focuses on validating safeguards through independent evals. AIUC publishes crosswalks to both.

How do I get AIUC-1 certified?

Through an accredited AIUC-1 auditor, who audits your controls, plus technical evaluations of your agents, currently conducted by AIUC; AIUC issues the certificate. Our AIUC-1 certification guide walks through each step.

What is AIUC-1?

AIUC-1 is an AI agent security standard covering security, safety and reliability. It is created by the Artificial Intelligence Underwriting Company (AIUC), which describes itself as a Public Benefit Corporation headquartered in San Francisco, and it is developed with the AIUC-1 Consortium.

This page summarizes what AIUC's own pages say about the standard, with links to each source. For the authoritative text, see the official AIUC-1 site.

Official source: aiuc-1.com

The AIUC-1 framework: six domains

The AIUC-1 standard groups its requirements into six areas:

A. Data & Privacy: data use policies, IP and trade secrets, PII and cross-customer exposure, secrets leakage.
B. Security: adversarial testing, input filtering, access controls, endpoint protection, secure code patterns.
C. Safety: risk taxonomy, pre-deployment and third-party testing, harmful and out-of-scope outputs.
D. Reliability: hallucination prevention and testing, restricting unsafe tool calls.
E. Accountability: failure plans, ownership, vendor due diligence, logging, transparency, quality management.
F. Society: preventing AI-enabled cyber misuse and catastrophic misuse.

How the AIUC-1 standard is updated

AIUC updates AIUC-1 each quarter, on January 15, April 15, July 15 and October 15, and documents every change in a public changelog. AIUC says certifications remain valid for the full year, with the standard version locked at certification.

AIUC-1 is AI-focused: AIUC says it does not cover non-AI risks that are addressed by frameworks or regulations such as SOC 2, ISO 27001 or GDPR.

The AIUC-1 consortium

AIUC says the AIUC-1 standard is developed and maintained with the AIUC-1 Consortium, which it describes as uniting 300+ security and risk leaders across the Fortune 500. All consortium members can propose changes each quarter, and proposed changes go through peer review before release.

AIUC's FAQ also names Technical Contributors from organizations including Orrick, Stanford, MIT, Carnegie Mellon, Cisco, MongoDB, Google Cloud and Faculty.

AIUC-1 and other frameworks

AIUC publishes crosswalks that map AIUC-1 to other frameworks and regulations, including SOC 2, ISO 42001, the EU AI Act, the NIST AI RMF, MITRE ATLAS and the OWASP Top 10.

In AIUC's words, ISO 42001 focuses on establishing AI governance frameworks and management systems, while AIUC-1 focuses on validating the robustness of safeguards through independent evals. SOC 2, AIUC notes, is not AI-specific and does not cover risks unique to AI.

AIUC-1 certification in brief

To earn AIUC-1 certification, an organization's AI agents must pass a control audit by an accredited auditor and technical evaluations, which AIUC says it currently conducts itself. The certificate itself is issued by AIUC. The full process is on our AIUC-1 certification guide, and the companies that have announced certification are on our list of AIUC-1 certified companies.

Want to see where your agent stands before an audit?

Take the 3-minute AIUC-1 readiness check

Frequently asked questions

What is AIUC-1?: AIUC-1 is a standard for AI agent security, safety and reliability, created by the Artificial Intelligence Underwriting Company (AIUC) and developed with the AIUC-1 Consortium. Its requirements cover six areas: data and privacy, security, safety, reliability, accountability and society.
Who created the AIUC-1 standard?: AIUC-1 is created by the Artificial Intelligence Underwriting Company (AIUC), which describes itself as a Public Benefit Corporation headquartered in San Francisco. AIUC says it develops and maintains the standard with the AIUC-1 Consortium of 300+ security and risk leaders.
How often is AIUC-1 updated?: Every quarter, on January 15, April 15, July 15 and October 15. AIUC documents each change in a public changelog.
How is AIUC-1 different from SOC 2 and ISO 42001?: AIUC says SOC 2 is not AI-specific and does not cover risks unique to AI, and that ISO 42001 focuses on AI governance and management systems while AIUC-1 focuses on validating safeguards through independent evals. AIUC publishes crosswalks to both.
How do I get AIUC-1 certified?: Through an accredited AIUC-1 auditor, who audits your controls, plus technical evaluations of your agents, currently conducted by AIUC; AIUC issues the certificate. Our AIUC-1 certification guide walks through each step.

Sources

AIUC-1: About AIUC-1. Accessed October 28, 2026.
AIUC-1: AIUC-1 Governance. Accessed October 28, 2026.
AIUC-1: Frequently asked questions. Accessed October 28, 2026.
AIUC-1 Consortium (aiuc-1.com). Accessed October 28, 2026.
AIUC-1 home page (aiuc-1.com). Accessed October 28, 2026.

Rends is not affiliated with or endorsed by the Artificial Intelligence Underwriting Company (AIUC), and is not an AIUC-1 auditor. The official source for AIUC-1 is aiuc-1.com.