Agent Safety and Governance Standards Accelerate
Key Questions
What initiatives are advancing agent safety standards?
Workday's Agent Passport tests against Mitre ATLAS with Cisco partnership, while OpenAI's Frontier Governance Framework aligns with EU AI Act and California law. A report shows only 11% of production agents pass security benchmarks, highlighting regulatory pressure.
How fragmented is health AI governance?
Health AI governance remains highly fragmented, with over 240 policies mapped across different frameworks. This fragmentation contrasts with accelerating standardization efforts in broader agent safety.
What is Microsoft contributing to agent governance?
Microsoft has proposed an Agent Control Specification (ACS) to provide runtime governance policies for AI agents. This adds to industry efforts like Workday and OpenAI initiatives for responsible deployment.
Why is there growing pressure for agent safety standards?
Multiple developments including security reports and regulatory alignments signal increasing industry and regulatory focus on responsible AI agent deployment. Only a small percentage of agents currently meet security requirements.
What does Veeam research reveal about rogue AI?
Veeam research examines accountability for rogue AI behavior, underscoring governance challenges. It complements broader calls for standardized safety frameworks across organizations.
Multiple initiatives push for standardized agent safety: Workday's Agent Passport (testing against Mitre ATLAS, Cisco partnership), OpenAI's Frontier Governance Framework (aligning with EU AI Act and California law), and a report finding only 11% of production agents pass security bar. Health AI governance remains fragmented with 240+ policies mapped. These developments signal growing industry and regulatory pressure for responsible AI deployment.