Safeguarding & technology
Powerful tools, kept on a short leash
Using technology with children raises fair questions, and we won't wave them away. This is how Foundry keeps a human responsible, constrains what the software can do, and stays honest about its limits.
Our commitments
What we hold ourselves to
A responsible human, always
Every Builder has a named Foundry Coach accountable for their learning and wellbeing. The technology assists that coach; it never replaces them.
No open-ended chatbot
Builders do not get a free-roaming chatbot to converse with. The software delivers structured lessons, practice and hints toward a defined objective, with no blank chat window that can be talked into doing the work for them.
Help separated from assessment
Hints and worked examples are kept apart from the controlled checks that award mastery, so support can never quietly inflate results.
Careful with Builder data
We collect what's needed to teach well and are clear about how it's used. See our Privacy page for specifics.
When the technology is wrong
What happens when the software makes a mistake
No technology is infallible, and we will not pretend otherwise. What matters is what happens when something is wrong. These are the safeguards the system is designed around, we do not describe any safeguard that is not actually in place.
Approved, reviewed content
Student-facing lessons, games, and challenges are drawn from approved curriculum content, not free-form generation in front of the child.
No open-ended chat box
There is no student chatbot. Adaptive intelligence chooses level, sequence, and difficulty in the background; the student sees a structured path.
Student flag and report
A student can flag content that looks wrong or confusing, which routes it to a human for review.
Coach escalation
If the academic system cannot resolve a legitimate issue, the Foundry Coach escalates it through the content and academic support process.
Separated from mastery
Practice, hints, and game points are kept separate from the controlled, hints-off checks that award mastery.
No blank chat box. No answer machine. No pretending technology is infallible.
Our promise about honesty
We will not claim a safeguard we don't have, describe a review process we don't run, or imply the technology is more capable, or more careful, than it is. If we discover we've gotten something wrong, our job is to fix it and tell you, not to hide it.
Questions about safety come first
If you have a safeguarding question we have not answered, include it with your pilot application and a real person will address it directly.