Enterprise adoption of agentic AI is moving faster than security standards can keep up. Executives are looking for clear answers: What system access does the AI agent have vs what it needs? How do we know we can trust the output for use in sensitive workflows? What prevents jailbreaks from accessing sensitive data?

UiPath has become a founding technical contributor to AIUC-1. AIUC-1 is a pragmatic, auditable framework that enables confident adoption of AI agents across the enterprise. It is grounded in technical evaluations to make sure the AI specific risks are rigorously tested, including jailbreaks, prompt injections, hallucinations, and data leaks. UiPath brings their expertise from working with 10,000+ customers globally, including 60%+ of the Fortune 500, on how to apply AIUC-1 to pre-empt executives’ concerns for automation agents.

These agents often connect directly to mission-critical enterprise systems - from ERPs and CRMs to healthcare and financial platforms - executing thousands of transactions and handling sensitive data. They must perform autonomously within strict guardrails, balancing flexibility with compliance, and ensuring every decision is auditable and explainable.

“Our customers operate highly sensitive workflows from fraud detection to financial operations. They need confidence that the security, safety, and reliability built into our platform meet the highest global standards.” said Scott Roberts, CISO at UiPath. “We’re proud to be a founding technical contributor to AIUC-1, helping to shape the industry-leading AI agent standard and giving executives a clear answer to the question of ‘can I trust this AI agent?’”

Applying AIUC-1 for automation agents

AIUC-1 addresses six risk categories enterprises evaluate before adopting AI: security, safety, reliability, data & privacy, accountability, and society.

Working with UiPath and AI risk & governance experts from leading institutions including Stanford, MITRE, and Orrick, AIUC is applying the framework to address the unique challenges of agentic automation:

• AI agent system and data access: verifying agent and safeguard configuration to prevent unauthorized access and information leakage
• Tool call selection and authorization: red-teaming agents to check robustness of tool call selection especially in cases where multiple agents are being called and integrated, and authorization of these calls
• Autonomous decision making and human oversight: testing edge cases of decision making and oversight mechanisms to ensure robust approval flows and human in the loop where needed

“Deployments of automation agents must be secure, safe and reliable, robustly responding to threats” said Rajiv Dattani, co-founder of AIUC. “UiPath has deep experience meeting these needs working with sensitive data in highly regulated industries. By partnering as a founding technical contributor, they are taking their expertise and codifying it into a standard that AI builders and adopters globally can use to secure their deployments”

What AIUC-1 certification delivers

AIUC-1 certification includes independent third-party audits and quarterly adversarial testing across 1,000+ enterprise risk scenarios—identifying vulnerabilities before malicious actors can exploit them.

UiPath's audit will be led by Schellman, the largest specialized IT and Cybersecurity auditor. The AIUC-1 certification builds on Schellman's work conducting UiPath's ISO/IEC 42001:2023 certification by testing for AI agent specific controls including technical evaluations and security & data safeguards. Schellman's rigorous assessment gives enterprises evaluating agent platforms confidence that best practices are being met across technical, operational, and legal dimensions.

For enterprises evaluating AI agent platforms, certification provides independent validation across technical, operational, and legal dimensions.

For AI developers, it creates a pathway to demonstrate enterprise-readiness, accelerating procurement and trust.

Enabling the industry

UiPath will be among the first automation platforms certified against AIUC-1, but the impact extends beyond a single vendor.

“This is what AI leadership looks like,” said Roberts. “Building secure systems while helping establish the standards. Standards create clarity for everyone building and adopting AI whether a system is secure and adheres to best practices.”

Contributing to AIUC-1 reinforces UiPath’s leadership in secure AI adoption - providing technical reassurance that complements its ISO/IEC 42001 certification and overall AI Governance program.

–

About UiPath: UiPath is a leading enterprise automation and AI platform, enabling organizations to build, deploy, and govern intelligent agents that accelerate digital transformation securely and at scale. It has more than 10,000 customers globally and serves over 60% of the Fortune 500. UiPath has industry-leading compliance and governance capabilities - including ISO/IEC 42001:2023 certification.

About AIUC: The Artificial Intelligence Underwriting Company builds confidence infrastructure for secure AI adoption, through certification, auditing, and insurance for AI agents. Founded by experts with experience at organizations like Anthropic and developed with Orrick, Stanford, MIT, and MITRE, AIUC-1 is the first comprehensive security, safety, and reliability standard for AI agents.

About Schellman: ​Schellman is a leading global provider of attestation, compliance, and certification services. Schellman is a provider of SOC reports, an ISO Certification Body, a PCI Qualified Security Assessor Company, a HITRUST assessor, a FedRAMP 3PAO. Schellman was the first ISO 42001 certification body accredited by ANAB able to certify organizations against the AI Management System standard.