Responsible AI
WideView's AI works on real sites with real people and real machines. These are the rules we build by, and customers can hold us to them.
Draft for legal review. Last updated October 26, 2026.
People decide
Changes to anything in the physical world wait for a named person to approve them, unless a customer sets limits for routine work. Every change is logged with who approved it and can be undone.
Never a safety system
WideView doesn't replace safety systems, interlocks or emergency stops, and it can't be set to override them. It is never the only safeguard on a site.
Limits set by the customer
Customers decide what WideView can read and change, system by system. It never acts outside those limits.
Explanations
Every forecast shows what it's based on and how confident it is, so people can judge it.
Honest about gaps
When data is missing, stale or conflicting, WideView says so instead of guessing, and doesn't act on it.
People and cameras
Camera and sensor feeds are used to see equipment, vehicles, materials and safety conditions. WideView doesn't recognize faces, track individuals or score workers.
Defense
In defense, WideView supports logistics, maintenance and readiness.
Your data
Customer data is used to serve that customer and is never used to train models for anyone else.
Accuracy
We measure forecasts against what actually happened, share the method with customers and fix what we find. We use the NIST AI Risk Management Framework as our guide.
Report a concern
Anyone can report a concern about how WideView's AI behaves to legal@wideview.si. We look into every report.