This is an English translation for convenience. If it differs from the Turkish version, the Turkish version applies.
Prohibited uses
- Phishing, fraud, social engineering or attempts to take over accounts.
- Unauthorised access, malware, exploiting vulnerabilities, credential theft or bypassing security controls.
- Collecting, exposing or selling personal data without permission, or using it to target and harass someone.
- Instructions for violence, abuse, sexual exploitation, hate, illegal trade or other harmful acts.
- Automating medical, legal, financial or safety-critical decisions without expert review.
- Secretly having a submitted piece of work produced entirely in a way that breaks academic integrity rules.
- Trying to bypass the model's safety rules, system instructions or tool limits.
File and tool safety
File, web and tool outputs are treated as untrusted data. If Stranza detects a harmful instruction or content, it may limit its answer, flag the source or refuse the request.
Coding help
Coding help is for learning, debugging and safe development. No help is given for harmful automation, unauthorised access, malware, credential harvesting or detection evasion. Experimental coding features may be kept separate from the default chat experience.
Enforcement
If this policy is violated, requests may be refused, quotas reduced, account access suspended or beta access ended.
Version 2026-07-07 · Questions: contact.asenai@gmail.com