AI helps you set things up. It never decides for your family.

AI runs when you set things up, not while your child learns

During setup, when you declare your values, your children’s learning styles, your schedule, and your approach, AI is actively helping interpret what you share. That is the right moment for it: you are providing input, and the system is learning from you.

When you press Build My Plan, or open your week, AI is not running again. A plain engine reads what you set and produces the result. The thinking happened up front, with you. After that, the system is simply following the configuration you gave it.

Same inputs, same result

An AI that gives a different answer every time is not a tool, it is an actor. A plan you build on Monday should not quietly shift by Friday when you have not touched anything.

So Blueprint’s outputs are deterministic. The same configuration produces the same plan, every time. If your plan changes, it is because you changed something, and you will always know why.

Your family decides

A computer cannot be held accountable for a decision, so it should never make one that belongs to you. Blueprint’s AI can prepare a draft, surface options, explain a tradeoff, or point out a gap in a plan.

It cannot commit a plan, change a child’s pace, or accept a family into a co-op on its own. Every decision that matters waits for a person to make it. That is not friction. It is where the authority belongs.

AI earns trust before it gets access

New AI capabilities do not start with access to your family. They earn it. A capability begins where no family sees its output, and moves closer to your children only after it has proven, over real use and human review, that it behaves consistently.

If a capability ever breaks that trust, it steps back and earns its way forward again. There is no shortcut that lets it keep going just because it already shipped.

You can turn it off

You decide how much AI touches each child, one child at a time, in that child’s settings. There are three choices:

• Intelligence off: no AI content for that child, anywhere.
• Blueprint Intelligence only: draws on what your family has declared, and nothing else.
• Blueprint + Anthropic AI: may also surface ideas from beyond your own data.

The setting lives on each child’s profile, and you can change it whenever you want.

Do not take our word for it

We would rather you verify than trust a claim. A few places to look:

• Transparency: ten things a family deserves to know, including how the matching works.
• Architecture of Refusal: exactly what we hold, what we do not, and how to leave.
• Our Principles: the commitments that govern every decision we make.
• Audits: our pre-registered checks on whether we are living up to all of it.