AI Game Studio

No. In AI Game Studio your teen builds a browser-based text adventure that uses AI as one component inside their own code. They write the story rules, the AI generates responses on command within those rules, and your teen stays the author and the programmer throughout. There is no open-ended companion to befriend or confide in. That kind of use — where a young person forms a relationship with a chatbot — is what Common Sense Media's risk assessments have rated as unacceptable for anyone under eighteen, because those products are built to create emotional attachment. Your teen treats the AI the way they would treat any other piece of their program: as something to direct and test.

Open-ended chat puts a teen on the receiving end of whatever the AI says. This course puts them on the other side of it. They decide what the AI is allowed to do, write the instructions that constrain it, and see exactly where it follows those instructions too literally or gets things wrong. A teen who has programmed an AI and watched it misbehave reads everything else it produces with a sharper eye. That shift — from consuming AI to directing it — is the whole point of the course.

No. The AI runs through a secure connection that Little Prodigies Academy sets up and manages. Students never log into an AI service or create their own accounts, and no personal details are sent to a chatbot company. They write and run their game code on their own computer, and the AI responses come back through LPA's controlled connection.

Play and share a working text adventure game they built from a starting template. Along the way they learn to write the instructions that tell an AI precisely how to behave (the part professionals call a system prompt) and to connect it to a program with a few lines of HTML and JavaScript. They also build the debugging habit of testing something, watching it break, and fixing it. Those skills carry well past a game: getting a machine to do what you actually meant, and the patience to improve something through its failures.

The game is powered by Claude, made by Anthropic, reached through LPA's controlled connection with strict usage limits and no open chat window. Sessions run live and stay capped at eight students, so an instructor sees what every teen is doing. Honesty matters here: no AI model is flawless, and that is exactly why the course teaches students to control the tool and question what it produces instead of treating it as an authority.