Is AI game making cheating?

That not writing the code means you did not make the game. Games have always been made by people who did not personally do every part. A composer who does not model, an artist who does not program, a designer who does neither — nobody has ever said those people did not make the game.

The same argument was made about every tool that removed a layer of manual work. Engines were cheating, because real programmers wrote their own. Visual editors were cheating. Asset stores were cheating. In each case the objection faded not because it was refuted but because the work moved somewhere else and everyone could see it.

And the tools were never neutral about who could participate. The requirement to learn programming first has kept a very large number of people with good ideas out of games entirely, and there is nothing noble about that filter.

That a lot of AI-made games are bad. This is true and worth conceding plainly. One-prompt games are the average of their genre, competent and anonymous, and there are a lot of them now.

But the reason they are bad is not that a model wrote the code. It is that nobody decided anything. A game with no decisions in it is bad whoever typed it, and hand-written games with no decisions in them have always been bad too.

The other fair version is about credit. Saying "I made this" while implying a kind of work you did not do is a claim about yourself, not about the tool. The fix is to be straightforward about how it was made, which costs nothing.

Deciding what the game is, and what it is not yet. Playing it and being honest about what is wrong. Making the calls of taste: exactly where the hill sits, how far the jump carries. Doing that repeatedly until it is right. That is most of the hours and all of the difficulty, and none of it moved.

What moved is the construction: the typing, the placing of every tree, and the requirement to acquire a programming skill before beginning. Those were real barriers and they were never the interesting part.

If anything the judgement load goes up. When implementing an idea costs a sentence, you test far more ideas, and someone has to decide which ones are good. That someone is you.

Some game jams restrict AI use, and some competitions and courses do too. There the rule exists and breaking it is straightforwardly cheating. Read the rules; they vary a lot and they change.

A course teaching you to program is a clear case. Having a model do the exercises defeats the entire point, and the person it costs is you.

Outside those, there is no rulebook for making a game. There is only whether it is good, and players have never asked how anything was made.

Flockbay asks before it builds: questions one round at a time, each with a recommendation, then a brief that says what gets built first, what comes after and what is not in this version. Nothing is built until you approve it. Then it builds the world, writes the GDScript and plays its own work to check it runs. You play the game and decide whether it is right, and when a call is pure taste, the same world tools are buttons in the Godot editor, so that one change can be your own hand. That is the argument above expressed as a product: the model does the construction, and the person makes every decision that gives the game its taste.

Being non-technical means you do not write the code. It does not mean you cannot judge the result, and the loop of playing and handing work back is the normal way to work here rather than a failure.

It is a desktop app for Windows, macOS and Linux with Godot inside. Your project is a normal Godot project, which is also the answer to anyone who wants to see how the game was made.