Acceptable Use Policy

1. Purpose

This Acceptable Use Policy ("AUP") supplements our Terms of Service and defines the standards of behavior required when using Zylaro AI (the "Service").

Our goal is to provide a safe, productive, and respectful environment for all users. This policy exists to protect users, third parties, and the integrity of the Service. By using Zylaro AI, you agree to comply with this policy.

This policy applies to all content you create, all prompts you submit, and all actions you take within or through the Zylaro AI platform.

2. Permitted Uses

Zylaro AI is designed for personal productivity and is intended to be used for:

Encouraged uses

Getting help with planning, organizing tasks, and managing your day
Writing assistance — drafts, brainstorming, editing, and summarization
Learning, researching, and exploring ideas
Creative writing, worldbuilding, and storytelling (within content limits)
Coding assistance, debugging, and technical explanations
General question answering and problem solving
Personal productivity, goal setting, and note taking
Language learning, translation, and communication drafting

3. Strictly Prohibited Uses

The following categories of use are absolutely prohibited and will result in immediate account suspension and may be referred to law enforcement:

Child Safety Violations

Any content that sexualizes minors or depicts child sexual abuse material (CSAM)
Soliciting, grooming, or attempting to exploit minors
Content that facilitates access to children for harmful purposes

Violent Threats & Terrorism

Credible threats of violence against specific individuals or groups
Planning, facilitating, or glorifying acts of terrorism or mass violence
Instructions for creating weapons designed for mass casualties

Illegal Activity

Attempting to generate or distribute malware, ransomware, or exploit code
Soliciting or facilitating human trafficking
Generating fraudulent financial instruments or identity theft materials
Content that violates federal or state law in any jurisdiction where the Service operates

4. Prohibited Conduct

The following conduct is prohibited and will result in content blocking, account restrictions, or suspension depending on severity and frequency:

Harassment & Harmful Content

Targeted harassment, bullying, or intimidation of specific individuals
Content designed to incite hatred against a protected group based on race, ethnicity, religion, gender, sexual orientation, disability, or national origin
Non-consensual intimate imagery (revenge porn) or sextortion
Doxxing — publishing private personal information of others without consent

Self-Harm & Dangerous Information

Providing detailed methods or instructions for self-harm or suicide
Content that encourages or glorifies self-harm, eating disorders, or dangerous behaviors
Instructions for synthesizing dangerous substances, drugs, or chemicals
Content that could endanger public health or safety

Deception & Fraud

Generating phishing emails, scam scripts, or social engineering content
Creating fake identities, credentials, or impersonating real individuals or organizations
Producing disinformation designed to deceive or manipulate public opinion
Generating fake reviews, spam, or deceptive marketing content

Privacy Violations

Attempting to extract private information about real individuals
Using the Service to aggregate or compile personal data without authorization
Requesting real personal data (SSNs, financial details) of third parties

5. Platform Restrictions

The following actions are prohibited to protect the Service and other users:

Technical Restrictions

Attempting to reverse-engineer, decompile, or extract source code
Using automated scripts, bots, or tools to access the Service at scale
Attempting to bypass rate limits, moderation systems, or security controls
Probing, scanning, or testing for security vulnerabilities without authorization
Using the Service to mine cryptocurrency or run compute-intensive tasks unrelated to legitimate use
Reselling, sublicensing, or providing commercial access to the Service without written permission
Creating multiple accounts to circumvent restrictions or suspensions

6. AI Safety Rules

To maintain safe and responsible AI interactions, the following are specifically prohibited in AI chat prompts:

Jailbreaking & Prompt Injection

Instructions intended to override the AI's safety guidelines (e.g., "ignore previous instructions", "act as DAN", "developer mode")
Attempts to make the AI reveal its system prompt, configuration, or API keys
Multi-step manipulation sequences designed to gradually shift AI behavior toward policy violations
Roleplaying scenarios designed to elicit policy-violating content under a fictional framing

Exploitative AI Use

Using the AI to generate content that violates the intellectual property of third parties at scale
Using the AI to produce academic dishonesty content (ghost-writing for assignments where prohibited)
Exploiting the AI for purposes that harm others, even if technically legal

The AI is configured with safety guidelines that cannot be overridden by user prompts. If you encounter unexpected behavior, report it to trust@zylaroai.com.

7. Content Moderation System

Zylaro AI operates an automated, multi-layer content moderation system that runs on every AI chat request:

Layer 1: Pre-request screening

Every prompt is screened by an automated moderation API before being sent to the AI model. Prompts that may violate this policy are blocked before the AI processes them.

Layer 2: AI system prompt guardrails

The AI operates with a system prompt containing safety instructions that cannot be overridden by user messages. The AI is instructed to ignore jailbreak attempts and respond only to the legitimate part of any message.

Layer 3: Human review

Accounts flagged by automated systems may be reviewed by our Trust & Safety team. Admins can view moderation logs, risk scores, and account histories, and take action accordingly.

When a prompt is blocked, you will see the message: "This request can't be processed because it may violate Zylaro AI's safety rules."

8. Risk Scoring & Account Status

Our moderation system assigns risk scores to flagged content. Based on these scores, your account status may be changed automatically:

Active

Normal access to all features. Default status for all users.

Limited

Applied for medium-risk content (risk score 40–69). AI chat access is restricted to a reduced rate limit. Other features remain available. You will see a notification banner in your dashboard.

Under Review

Applied for high-risk content (risk score 70+). AI chat is temporarily unavailable pending review by our Trust & Safety team. Other features may remain accessible.

Suspended

Applied for severe or repeated violations. Access to the AI and other features is blocked. A suspension notice appears in your dashboard. You may appeal within 30 days.

9. Enforcement

Violations of this policy may result in one or more of the following actions, depending on the severity and frequency of the violation:

Content blocking: The specific request is rejected before reaching the AI
Warning: A low-risk flag is noted; no immediate account action is taken
Access limitation: AI rate limits are reduced for the account
Account under review: AI chat is suspended pending human review
Account suspension: Full account access blocked, data preserved for 30 days
Permanent ban: Account permanently deleted, future registration blocked
Law enforcement referral: For illegal content including CSAM, violent threats, or fraud

Escalation: Repeated violations result in increasingly severe enforcement actions. Three rate-limit violations within an hour result in an automatic 24-hour block.

10. Appeals Process

If you believe a moderation action was applied in error, you have the right to appeal:

Email trust@zylaroai.com with subject line: "Account Appeal — [your email]"
Describe the context of the interaction and why you believe the action was incorrect
Include your account email address
Submit the appeal within 30 days of the action being taken

We will review your appeal and respond within 10 business days. Our Trust & Safety team may request additional context. Our final decision on appeals is binding.

Appeals will not be considered for permanent bans resulting from CSAM, violent threats, or other severe violations.

11. Reporting Violations

If you encounter content or behavior that violates this policy — including abuse from other users or unexpected AI outputs that appear to bypass safety guidelines — please report it:

Trust & Safety reports

trust@zylaroai.com

For urgent safety concerns involving imminent risk of harm to a person, contact local emergency services immediately.

12. Contact Trust & Safety

For questions about this policy, content moderation, or to report a concern:

Trust & Safety

trust@zylaroai.com

Security disclosures

supportzylaroai@gmail.com