Trustworthy & Responsible AI
We use AI to assist, not replace. Our practices align with DHS trustworthy AI principles, NIST AI RMF 1.0, and OMB M-24-10 safeguards.
Quick Summary
SecurePoint USA uses AI to assist with entity matching and risk summarization during visitor screening.Humans make all final access decisions, and every action is logged in immutable audit logs. Our AI practices are built on a foundation of safety, security, and transparency.
Methods
- Entity matching with fuzzy/phonetic search
- Risk summarization to highlight context
- Optional vision parsing for IDs (opt-in)
Safeguards
- Human-in-the-loop adjudication
- Immutable audit logs for actions and outcomes
- Evaluator thresholds and bias checks
- Policy templates for ITAR/EAR and sanctions
Frequently Asked Questions
Is AI required to make access decisions?
No. AI assists. Final access decisions are made by authorized personnel.
How are AI outputs governed?
We align with NIST AI RMF, apply evaluator thresholds, and maintain immutable audit logs.
Do you support DHS trustworthy AI principles?
Yes. We implement human-centered safeguards, transparency, and monitoring consistent with DHS guidance.
Is ID vision parsing mandatory?
No. Vision parsing is optional and disabled by default; organizations may opt in.
Important Disclaimer
AI outputs are assistive signals, not determinations. Compliance and access decisions are solely made by authorized personnel.