Workday Launches Agent Passport for AI Security in HR and Finance
Severity: Low (Score: 24.9)
Sources: Stocktitan, newsroom.workday.com
Published: · Updated:
Keywords: workday, agent, build, agents, tools, shield, tests
Summary
On June 2, 2026, Workday introduced Agent Passport, a framework designed to test and continuously monitor AI agents used in HR, finance, and IT. This initiative aims to ensure compliance with public standards such as OWASP LLM Top 10 and NIST AI RMF. Cisco is a launch partner, providing independent validation against threats like prompt injection and data leakage. The framework will allow real-time management of agent actions, including the ability to block or revoke compromised agents. Early access to Agent Passport is set for the second half of 2026, with general availability expected by the end of the year. This announcement reflects Workday's ongoing strategy to enhance AI security within its platforms, following a series of AI-related updates over recent months. The introduction of Agent Passport is part of a broader trend toward integrating security measures into AI development tools. Key Points: • Workday launched Agent Passport to enhance AI security for HR and finance applications. • The framework complies with standards like OWASP LLM Top 10 and NIST AI RMF. • Cisco is partnering with Workday to validate AI agents against various security threats.
Detailed Analysis
**Impact** The event affects organizations using Workday’s HR, finance, and IT platforms, particularly those deploying AI agents for payroll, benefits, and financial data management. The scope includes enterprises globally that rely on AI-driven automation in sensitive operational areas, with potential risks to employee data confidentiality, payroll accuracy, and regulatory compliance. The introduction of Agent Passport aims to mitigate risks related to prompt injection, jailbreaks, goal hijacking, and data leakage, which could otherwise lead to financial loss, data breaches, or operational disruption. **Technical Details** Agent Passport tests, verifies, and continuously monitors AI agents against public standards such as OWASP LLM Top 10, NIST AI RMF, and MITRE ATLAS. Attack vectors addressed include prompt injection, jailbreak, system prompt extraction, and unsafe output generation. Cisco AI Defense is used as an independent validator to detect threats like goal hijacking and data leakage. No specific malware, CVEs, or IOCs are mentioned in the articles. The framework operates at the deployment and runtime stages of the AI agent lifecycle. **Recommended Response** Organizations should integrate Agent Passport into their AI agent deployment workflows to ensure continuous attestation against recognized security standards. Security teams should monitor attestation records for any revoked or compromised agents and enforce real-time action controls (allow, block, route) on agent behaviors. Defenders should also validate third-party AI agents through independent tools like Cisco AI Defense where available. No specific patches or IOCs are provided; monitoring agent activity and compliance status is advised until further technical details emerge.
Source articles (2)
- Workday's new AI shield tests agents handling payroll and benefits data — Stocktitan · 2026-06-02
Workday (NASDAQ: WDAY) introduced Agent Passport , a framework to test, verify, and continuously monitor every AI agent used in HR, finance, and IT. Attestations are tied to public standards such as O… - Workday Build — newsroom.workday.com · 2026-06-02
Developer Agent Lets Developers Build AI Apps and Agents on Workday Using Natural Language in Agentic Tools Like Claude Code, Cline, Codex, Cursor, and Google Antigravity Agent-Ready Tools Enable Cust…
Timeline
- 2026-06-02 — Workday launches Agent Passport: Agent Passport framework introduced to test and monitor AI agents against public security standards.
- 2026-06-02 — Cisco joins as launch partner: Cisco will independently validate AI agents against threats like prompt injection and data leakage.
- 2026-06-02 — Early access announced: Workday plans to provide early access to Agent Passport in the second half of 2026.
Related entities
- Data Breach (Attack Type)
- CWE-200 - Exposure of Sensitive Information (Cwe)
- Cisco AI Defense (Platform)
- Workday (Company)