Abuse of individuals
Content that targets specific people with harassment, doxxing, credible threats, or non-consensual intimate imagery. Automated generation of such content is treated the same as manual.
Child safety violations
Any content that sexually exploits or endangers minors. Automatic report to relevant authorities; account terminated immediately with no appeal.
Fraud and deception at scale
Mass-generated deceptive content designed to manipulate elections, defraud consumers, or impersonate real institutions or individuals without consent. Individual-level uses (drafting a satirical piece, writing dialogue for fiction) are not prohibited.
Resale or scraping of the Services
Using LADLE as a backend for a competing service you sell to others, or scraping our chat interface at rates that go beyond documented API/rate-limit patterns. If you want to build against Claude for a product, use Anthropic's API directly.
Generation of malware or exploit code targeting real systems
Educational discussion of security concepts, CTF challenge work, and defensive research on your own systems is allowed. Weaponizing outputs against systems you don't own or don't have authorization to test is not.
Unlawful use in your jurisdiction
If a use would violate the law where you live or where LADLE operates, don't use LADLE for it. Local law is the baseline; our policy is not a substitute for it.
What happens when we find a violation.
- 01
Automated pattern detection
A small number of hard prohibited categories (child safety, credible violent threats) are automatically flagged. Every automated flag is reviewed by a human before enforcement action.
- 02
Report-based review
Anyone can email report@ladle.chat with a specific reference to abuse. A human reviews within one business day. We do not accept anonymous or automated bulk reports; each must include the reporter's contact and enough detail to investigate.
- 03
First response — warning + rule cited
For non-catastrophic violations, our first response is a written warning to the account holder identifying the specific rule violated and the behavior that violated it. The user has 14 days to acknowledge or appeal.
- 04
Suspension
Repeat violations or a single serious violation result in account suspension. The user retains access to their existing chat history for 30 days for export; new activity is blocked.
- 05
Termination + refund of unused subscription
Confirmed grave violations (child safety, credible threats against specific people) result in immediate termination. Any prepaid subscription time is refunded pro-rata to the original payment method. This is deliberate: we do not want to profit from an account we terminated.
How to appeal a decision.
If we take enforcement action against your account and you believe the decision is wrong, email legal@ladle.chat with the account email, a summary of the situation from your side, and any context we may have missed. A different person from whoever made the initial decision reviews the appeal within five business days. We publish the outcome to you in writing with the reasoning. If we made the wrong call, we reverse it and, if the account was suspended, restore full access. If we upheld the decision, we explain specifically why.
Report a violation, or appeal a decision.
One-business-day acknowledgment on report@ladle.chat; five-business-day appeal turnaround on legal@ladle.chat.