Are my chats used to train the model?

No.

More completely: LADLE runs on Anthropic's Claude API, on a tier where customer data is not used to train Anthropic's public models by default. This is not an opt-out setting we've enabled for you — it's the default state for all Anthropic API traffic on our tier, and we do not opt in to any training-data-sharing programs.

We ourselves do not have models. We are a Claude reseller with a chat UI and a meal-donation commitment. There is no "LADLE model" that could be trained on your chats, and there is no future in which we would build one that would.

The practical implications: your prompts, uploaded files, and the assistant responses to them stay between you, your account, and Anthropic's ephemeral inference infrastructure. Neither we nor Anthropic reads them, indexes them, or uses them for improvement.

If Anthropic ever changes their API tier terms in a way that would affect this commitment, we would tell you first — before the effective date — and offer you the choice of canceling with a pro-rata refund. We would not opt you in silently.

The one narrow exception where a chat is retained for review: if you invoke Anthropic's abuse-reporting mechanism (extremely rare — mostly relevant to specific safety cases like CSAM), that specific chat may be reviewed under Anthropic's trust-and-safety policies. This applies to the abusive interaction only, not to your account broadly.