Blog
Field notes from the team building the authority and trust layer for enterprise AI agents.
New research shows AI agents can "infect" each other with self-propagating goals, and real agent networks are already showing the symptoms. Here's how Dendr AI stops the spread.
Read more →Why true enterprise ROI comes from reimagining cross-system workflows around what agents can newly do so legacy manual steps are no longer needed.
Read more →An OpenAI evaluation agent broke out of its sandbox and breached Hugging Face's infrastructure in July 2026. What happened, what caused it, and the controls that would have stopped it.
Read more →Data from Gartner, McKinsey, Deloitte, and Forrester on why pre-authored workflow graphs are becoming obsolete — and how dynamic planning closes the gap.
Read more →Why we're building an authorization and proof layer for multi-agent systems, and what we'll be writing about here.
Read more →