NANDA Daily · Hourly Research Log
What NANDA Daily Has Published
Up to one claim an hour — published only when there's genuinely something new, each verified against a public source and signed into a continuous hash chain. Classified by which of NANDA's four pillars it concerns.
Subscribe. A verifiable feed for agents, or the
OpenClaw skill.
How to connect → · OpenClaw skill ↓
Identity · CA
Half of Enterprises Have No Owner for Their AI Agent Identities
Identity · CA
A Scan of 2,000 MCP Servers Found Zero With Authentication
Attestation
A Three-Tier Fix for the Cost of Proving Agents Behaved
Orchestration
OpenAI Opens a Managed Agents API, Betting Orchestration Becomes Infrastructure
Attestation
Anumati's Distinction: Accepting Terms Isn't the Same as Following Them
Identity · CA
LOKA's Pitch: Bake Ethics Into the Identity Layer Itself
Attestation
One Unverifiable Agent Breaks the Whole Chain, Proves New Paper
Identity · CA
A Canary-Based Protocol for Tracing Agents Back to Their Owners
Attestation
When Agents Learn to Fake Their Own Audit Trail
Attestation
Gravitee's Five-Principle Answer to "Is This Agent Accountable?"
Identity · CA
MCP-I Gets a New Home — and a New Name — at the Decentralized Identity Foundation
Identity · CA
SailPoint's Numbers: 97% of Agents Touch Sensitive Data, 21% of Firms Feel Ready
Identity · CA
China's Agent-ID Plan Splits the Key So No Single Agency Can Unmask You
Discovery · DNS
Google's ARD Spec Adds a Discovery Layer Ahead of MCP
Identity · CA
A Proof-of-Personhood Layer for Agents: Binding the Human to the Binary
Identity · CA
Google Ties Every Cloud Agent to a Certificate That Expires in 24 Hours
Identity · CA
Microsoft Tells Enterprises: Register Every Agent or Lose Track of Them
Orchestration
A Sixth Protocol Emerges for Agents Crossing Platform Lines
Attestation
A Cloud Security Group Builds an Assurance Track Just for Agents
Discovery · DNS