Are our apps vulnerable to the WeWeb 3.0 AI?

This might be a bit naive on my part, but does the WeWeb AI use our project data to continually train the model?

Is there any mechanism to prevent my app from being copied by the AI? The scenario I’m thinking about is one of my larger users that has 8 subscriptions decides they want to build it themselves. They use DevTools and see the app is built on WeWeb. Could they create a WeWeb account and ask the AI to duplicate my specific pages? I understand they would still need to figure out the backend, but Xano will have AI assisted function stack creation sooner than later.

fwiw, I’m less worried about the look/design, and more worried about my workflows which I would consider intellectual property as they are unique to the niche my app serves.

This is improbable. Also, WeWeb doesn’t use their own model. So there is nothing to be trained. I don’t think there is anything to be worried about.

1 Like

No.

1 Like

mmh yes and no i would say, what i mean is that set it as blackboxed by default without much control by weweb and therefore some guardrail if the openai or claude (didn’t check the model used) would happen to have some kind of leak or just a terms and conditions allowing them to trained on output/inputs call
even if the provider has a process of anonymization i wouldn’t set a probability of 0 or 100 tbh

and don’t think it’s a vanilla foundation model, imo most likely it’s a prompted/pre-trained version of a foundation model with a given knowledge base fed.

and there is always something to be trained on if they want in theory (don’t say they do)