sentinelK
139 天前
anthropic 是为数不多的明确表达用户数据不会被直接训练成“Claude”模型,所以他也是最贵的之一。
虽然但是,他仍然给自己留了口子:your conversations are flagged for Trust & Safety review (in which case we may use or analyze them to improve our ability to detect and enforce our Usage Policy, including training models for use by our Trust and Safety team, consistent with Anthropic’s safety mission)(只要是他们认为“触发机制”的,他们就可以存储)
以及:We will not use your Inputs or Outputs to train our generative models (i.e. Claude)(不能训练“ our generative models”,这个 our ,就很耐人寻味)