🧠 Responsible AI Usage Policy

Siyarix experiments with Artificial Intelligence to plan tasks and assist operators. Because AI is unpredictable, this policy outlines how I try to handle AI responsibly within the platform.

🏗️ How We Use AI

1. Task Planning: Trying to convert natural language goals into execution plans.
2. Interactive Chat: Providing an assistant interface.

👁️ Human Oversight is Mandatory

AI is an assistant, not an autopilot. - The Permission Gate: Every command is checked before execution. - User Confirmation: If the AI suggests a dangerous command, execution halts, and you must explicitly approve it.

Danger

You are legally responsible for the commands you authorize. Never blindly approve a command you do not fully understand.

⚠️ Known Limitations of AI

Current AI models can be flawed. Be aware that the AI may: - Generate incorrect execution plans. - Suggest tools that aren't installed. - Misinterpret your intent. - Hallucinate fake flags or IP addresses.

🛡️ Data Protection & Privacy

When you use cloud-based AI providers (like OpenAI or Gemini), data leaves your machine. - Siyarix uses a lightweight DLP engine to try to mask basic credentials before they hit the internet. - For maximum privacy, Siyarix supports local, offline models (via Ollama or LM Studio) so your data never leaves your laptop.

🚩 Reporting AI Misbehavior

If the AI suggests something harmful or hallucinates dangerously: - Please open an issue on the GitHub Repository. - Do not include actual credentials in your bug report!

⚖️ Your Ethical Obligations

By using Siyarix's AI features, you agree to: - Never use the AI to generate targeted phishing campaigns. - Never prompt the AI to write destructive malware. - Never attack systems outside your authorized scope.