Just got the updated terms notification from Writesonic. Section 4.2 is a major red flag on training data.
They now claim broad rights to use customer inputs/outputs to "improve and train" their models. No explicit opt-out mentioned.
* This is a cost trap waiting to happen.
* Feeding proprietary data into their model? That's a massive liability.
* If your prompts contain any internal data, you're potentially giving it away.
Has anyone with a legal or compliance background reviewed this? Specifically on:
* Data ownership post-submission
* How this aligns with GDPR/other regs
* If there's a way to disable this in account settings (I couldn't find one)
show me the bill
Yeah, that clause caught my eye too. It's becoming a pattern with these "all-in-one" content generators. Not a lawyer, but our compliance team freaks out over this exact thing.
They might argue it falls under "legitimate interest" for GDPR, but that's shaky if you're putting client briefs or product roadmaps in there. The data ownership angle is the real kicker - you own it until you submit, then the license grant kicks in.
Have you checked if they offer a Business/Enterprise tier? Sometimes the data opt-out is gated behind that paywall, which is a nasty practice. If it's not in settings, you probably have to email them.