Legal
Acceptable Use Policy
This Acceptable Use Policy (“AUP”) describes prohibited activities when using KeyForge. It is incorporated into our Terms of Service.
Effective date: July 6, 2026·Last updated: July 6, 2026
1. Your responsibility
You are responsible for all activity conducted through your account, including by your applications and autonomous agents. You must ensure that your use of KeyForge and any downstream model providers complies with this AUP, our Terms of Service, applicable law, and the usage policies of the model providers your requests reach.
2. Prohibited uses
You may not use the Service to:
- Violate any law or regulation, or infringe the intellectual property, privacy, or other rights of others.
- Generate or distribute malware, facilitate intrusion, credential theft, phishing, or other malicious activity.
- Circumvent, probe, or overload the Service or model providers, including through scraping, denial-of-service, or evading rate limits, budgets, or virtual-key controls.
- Violate the acceptable-use or content policies of any downstream model or infrastructure provider.
- Generate content that sexually exploits or endangers minors, incites violence, promotes self-harm, or facilitates the creation of weapons or other clearly harmful outcomes prohibited by law or provider policy.
- Engage in unlawful surveillance, harassment, discrimination, or generation of unlawful deceptive content.
- Misrepresent AI-generated content as human-authored where doing so is deceptive or unlawful, or use output to deceive in ways that cause harm.
- Resell, sublicense, or expose raw provider credentials or the Service in a way that violates these policies.
3. Autonomous-agent safeguards
Where your agents can take consequential actions — such as making purchases, sending communications, or modifying systems — you must implement appropriate human oversight, testing, and guardrails. Do not deploy agents in ways that create unreasonable risk to people, property, or third-party systems without adequate controls. You are responsible for validating model output before acting on it.
4. Enforcement
We may investigate suspected violations and take action, including throttling, suspending, or terminating access, revoking virtual keys, and removing offending content or configurations. Where practical and lawful we will provide notice and an opportunity to cure, but we may act immediately to protect the Service, other customers, providers, or the public.
5. Reporting abuse
To report a violation of this AUP, contact [email protected]. Metis reviews reports and responds as appropriate.