Acceptable Use Policy
Effective August 13, 2026 · SHA-256 6d878a7b23ba3663c28cb2047ba97661f82d760c22689dedf454070ef9b2d499
This Acceptable Use Policy applies to every Farka account, dataset, model, run, export, deployment, and API request. Model-provider policies may impose additional restrictions.
Lawful rights
Do not upload or process data, code, weights, personal information, confidential material, or intellectual property without the rights and lawful basis required for the requested use.
Safety and harm
Do not use Farka to facilitate violence, exploitation, abuse, unlawful discrimination, harassment, stalking, self-harm encouragement, dangerous wrongdoing, or evasion of safety controls.
Privacy and identity
Do not perform unlawful surveillance, biometric identification, identity theft, impersonation, credential theft, or processing of sensitive personal data without documented authority and safeguards.
Security
Do not distribute malware, exploit systems without authorization, disrupt service, probe other tenants, bypass access controls, expose secrets, or use generated artifacts for malicious cyber activity.
Regulated and high-impact use
Do not rely on model output as the sole basis for medical, legal, financial, employment, housing, education, insurance, credit, law-enforcement, migration, or other high-impact decisions. Use qualified human review and applicable safeguards.
Content integrity
Do not generate unlawful deceptive content, non-consensual intimate content, child sexual abuse material, or undisclosed synthetic media where disclosure is legally required.
Platform integrity
Do not resell access contrary to your plan, evade billing or quotas, overload providers, automate account creation, scrape protected data, or misrepresent Farka verification and benchmark evidence.
Enforcement and contact
We may block a quote, suspend a run, preserve evidence, restrict an account, or report activity where required by law. Appeals and questions may be sent to edihasaj@gmail.com.