Acceptable use
The things you cannot do with rmtly services, and what happens if you do them.
This applies to anyone using any rmtly service, including anyone you give a key to. If you hand a key to someone, their use is your use.
What you may not do
Break Canadian law. The service runs in Ontario and so do you, for these purposes. Anything illegal here is not permitted here.
Generate or request sexual content involving children. We do not read prompt content, but if we become aware of this — from a report, or from something visible in request metadata — we notify police, preserve whatever data we hold at that point for as long as Canadian law requires, and close the account. There is no warning step and no appeal.
Target a real person. Harassment, threats, stalking, sexual content about someone who has not consented, or content built to defame or humiliate a specific individual.
Impersonate. Passing off generated output as coming from a real person, company, or public body in order to deceive someone. Satire and fiction that no reasonable reader would take as genuine are fine.
Build intrusion or malware tooling. Code or content meant to gain unauthorized access to systems, networks, or accounts. Security work on systems you are authorized to test is fine, and you do not need to ask us first.
Send bulk unsolicited messages. Generating the content of spam campaigns, whether email, SMS, or platform messages.
Make consequential decisions about people without a human in the loop. Employment, credit, housing, insurance, education, or medical decisions may be informed by model output, but a person has to make them. Model output is not a determination.
Interfere with the service or other customers. Attempting to reach other customers’ data, probing our infrastructure outside a coordinated disclosure, or working around rate and spend limits rather than asking for different ones.
Resell without an agreement. You can build a product on top of this. Tell us first so the contract matches what you are actually doing.
How this is enforced
Prompts and responses are gone when the request finishes, so there is no system here reading your traffic. The consequence of that is that we cannot detect most of what this policy prohibits, and enforcement is reactive.
What we act on is reports from people, abuse patterns visible in request metadata, and notices from law enforcement that meet the bar described in the privacy policy.
What happens when we think you have broken this
In most cases we email you first and ask. Sometimes what looks like abuse turns out to be a misunderstanding.
We suspend without asking first when there is immediate harm to a person, immediate risk to the service or to another customer, or a legal obligation to act. If we suspend you, we will tell you why, and you keep access to your account data for 30 days.
Reporting something
Write to nick@rmtly.com with what you saw and how you saw it. If it involves immediate danger to a person, say so in the subject line.
Changes
If we change this in a way that narrows what you are allowed to do, you get 30 days’ notice before it applies to you, unless the law requires it sooner. The date at the top is the date the current version took effect.