- Cost
- Free, no obligation
- Reply window
- ≤ 2 business days
- Offices
- Seattle · Singapore
WTIA AI Summit · Seattle · October 2026
A free risk scan for the agent you are about to trust.
Human Intelligence publishes HI-AAF, the Agent Assurance Framework for AI agents in production, assesses agents against it, and operates the human review layer that keeps the evidence current after deployment. If you run an AI agent, or are about to, describe it below and we will read it against the framework. There is no charge and nothing to install.
§ I · What we do
- The Standard
- HI-AAF v1.3: nine domains and ninety-eight controls for AI agents in production, mapped to NIST AI RMF, ISO/IEC 42001, OWASP LLM Top 10, MITRE ATLAS, SOC 2, the EU AI Act, and Singapore’s MAIGF. Open for public review. Read it.
- The Practice
- Workshop, Readiness Assessment, Assessment, Maintenance, and Certification, carried out by Human Intelligence assessors and concluding, where the agent meets the Standard, in a written assessment. The engagements.
- The human review layer
- Trained human reviewers handle what automated checks cannot decide, so the evidence behind an agent stays current after it is deployed rather than on the day it was assessed. How it operates.
§ II · The scan
What it is
You describe one agent: what it does, which systems and tools it can act on, and who it acts for. A Human Intelligence principal reads that description against the nine HI-AAF v1.3 domains and returns a short written note of the most material risks we can see, each tied to the domain it falls under and the question to put to your team.
Nothing is installed and no access to your systems is needed. The scan works from what you tell us.
What it is not
The scan is not an Assessment and does not produce a Letter of Assessment. It is a reading of what you tell us, free and without obligation.
If it raises questions worth a longer conversation, the next step is a Workshop.
§ III · Request
HI-AAF v1.3 is an independent framework published by Human Intelligence. It is not affiliated with NIST, ISO, AICPA, or any government or accreditation body. The framework is offered openly for review during the current draft cycle.