- You can always stop it. A stop that works mid-task, offline, and whatever the AI is doing. No update, remote command or AI output can override it. Directive 1
- You set the limits. It does only what it has declared, and asks before anything that matters: spending money, messaging people, sharing your data, unlocking doors. Directive 2
- You know what you are dealing with. It says it is AI, never pretends to be a person or a professional, and labels the media it makes. Directive 4
- Someone answers for it. A named, reachable company, incidents reported within 72 hours, and logs that show what happened and on whose instruction. Directive 9
The principle
Human in Charge
“Human in charge” means a person can always stop an AI product or robot, decides what it may do on its own, knows when they are dealing with a machine, and can hold a real company to account. It is the idea behind every Humasure Directive.
In plain words
Four things that keep a person in charge
Related ideas
In the loop, on the loop, in charge
People use several phrases for keeping humans involved in AI. They are not the same thing.
| Phrase | What it usually means | What it can miss |
|---|---|---|
| Human in the loop | A person approves each decision or step before the system acts | Does not say who can stop it, or what happens when no one is watching |
| Human on the loop | A person monitors the system and can step in | Monitoring is not control if the stop can be delayed or overridden |
| Human in charge | A person can always stop it, sets what it may do alone, knows it is a machine, and a company is accountable | Only means something if it is tested. That is what certification is for |
Why now
The machines are moving into our homes
AI agents can now spend money and send messages for us. Companion chatbots talk to children and lonely adults for hours. Home robots that see, hear and move around the house are going on sale. Each of these is useful, and each raises the same question: when it matters, is a person still in charge?
Laws are arriving piece by piece and differ from state to state and country to country. Buyers can’t test a robot’s stop button or a chatbot’s honesty in the store. An independent mark, backed by testing anyone can look up, lets them check in seconds.
A principle for everyone. “Human in charge” is not a trademark, and we hope every company, regulator and researcher uses it. Humasure™ is our name, and the Humasure mark is how a product shows it has been independently tested to keep a human in charge.
Help set the standard
We are recruiting founding advisors and our first ten products in AI companions, eldercare and home robots. Tell us who you are and how you’d like to take part.
- ProducersJoin the Founding Cohort and certify at cost in exchange for a public case study.
- ExpertsRobot safety, AI security, child safety, consumer protection: review the draft or join the Standards Council.
- Families and organizationsComment on the Directives during the public comment period.