AI Safety & Content Safety Policy

Last updated: 15.11.2025

At Spayo (“we”, “our”, “us”), we are committed to developing and delivering AI-powered design tools that are safe, transparent, and aligned with global best practices in responsible AI. This AI Safety & Content Safety Policy explains how we ensure that Spayo’s AI systems operate ethically, minimize risks, and protect the rights, safety, and interests of our users and the public.

Our goal is simple: Creativity powered by AI — without compromising trust.

1. Scope of This Policy

This policy applies to:

  • app.spayo.xyz
  • www.spayo.xyz
  • all AI models, prompts, pipelines, and automated processes used within the Spayo application
  • all generated designs, images, transformations, and AI outputs
  • all user-submitted content (“User Content”)

This policy supplements our Terms of Service, Privacy Policy, and Acceptable Use Policy (AUP).

2. Our AI Safety Principles

We follow five core principles:

2.1. Safety

We design our system to prevent harmful, dangerous, or illegal outcomes. This includes filtering, monitoring, and restricting unsafe content.

2.2. Transparency

We clearly communicate how our AI works and what its limitations are. Users always know when they interact with AI.

2.3. User Control

Users maintain control over their uploaded content, data, and decisions. They can delete their data or opt out of processing at any time.

2.4. Privacy & Security

We do not train on user-uploaded personal data. We apply strict access controls, encryption, and minimal data retention.

2.5. Fairness & Non-discrimination

Our service is monitored and tested to reduce biased or unfair output.

3. What Our AI Does

Spayo’s AI is designed exclusively for visual redesign and creative transformation of images, such as:

  • Interior design transformations
  • Exterior redesign
  • Automotive redesign
  • Yard, garden, and landscape redesign
  • Style transformations using predefined themes
  • Custom prompt–based aesthetic changes
  • Color, material, and atmosphere modifications

Our system does not:

  • Identify or recognize individuals
  • Perform biometric processing (e.g., face recognition)
  • Diagnose health conditions
  • Predict personal attributes
  • Generate deepfakes of real people

Spayo is classified as a “minimal risk AI system” under the EU AI Act definitions.

4. User Content Restrictions (Safety Filters)

To maintain safety, Spayo does not allow uploads or prompts that contain:

4.1. Illegal or Dangerous Content

  • Weapons, violence, or criminal activity
  • Sexual content, pornography, minors in risky contexts
  • Content promoting self-harm
  • Explicit hate or extremist imagery

4.2. Personal or Sensitive Identifiable Data

  • Driver’s licenses, passports, ID cards
  • Financial data
  • Medical images
  • Biometric data
  • Photos of other people without consent

4.3. Copyright Violations

  • Uploading copyrighted media you do not own
  • Asking the AI to recreate trademarked or proprietary designs (e.g., “Redesign this car like a Porsche GT3 exact replica”)

Violations may result in blocking the request or restricting account access (see Acceptable Use Policy).

5. Safety Mechanisms in Our AI System

5.1. Upload Safety Scanning

Before processing, all uploaded images are screened for:

  • Nudity
  • Violence
  • Inappropriate content
  • Explicit or unsafe imagery
  • Personally identifiable documents

Unsafe images are automatically rejected.

5.2. Prompt Safety Filtering

We analyze user-provided prompts using safety classifiers to block harmful requests.

5.3. Output Monitoring

We perform human and automated evaluation to detect biased, unsafe, or unintended model behavior.

5.4. Rate Limiting & Abuse Prevention

Systems are in place to prevent automated misuse, bulk exploitation, or malicious generation attempts.

6. How We Use Data for AI

✔ We do not use user images to train AI models.

✔ We do not sell personal data.

✔ We do not share personal data with advertisers.

✔ We store only what is required for account and service functionality.

We may use aggregated, anonymized technical logs (never personal data) for system improvement.

Examples of safe data we may analyze:

  • Model performance metrics
  • Error logs
  • Non-identifying usage statistics
  • Style selection frequencies

7. User Rights Related to AI Decisions

You have the right to:

7.1. Request human review

If you believe an upload or prompt was incorrectly blocked, you may request manual review.

7.2. Opt-out of data processing

You can request deletion or export of your data at any time via:

➡ “Delete or Access My Data” Page

7.3. Transparency Requests

You can ask:

  • How our AI models work (high-level explanation)
  • What data categories are processed
  • How AI safety decisions are made

8. Limitations & AI Disclaimers

While we strive for high quality, AI-generated content has inherent limitations:

  • AI may produce unexpected or imperfect outputs
  • Generated content may contain inaccuracies
  • Designs should not be relied on as professional architectural, engineering, or safety advice
  • Output may vary depending on input quality and conditions

Users are responsible for reviewing generated results before applying them in real-world projects.

9. Reporting Harmful Content

If you encounter:

  • Unsafe output
  • Biased or discriminatory content
  • Copyright issues
  • Privacy concerns
  • Technical errors

Please contact us immediately:

Email: [email protected]
Contact Form: https://www.spayo.xyz/about/contact

We investigate and respond to all reports.

10. Model Updates & Continuous Improvement

We regularly:

  • Update safety filters
  • Improve output quality
  • Audit system behavior
  • Retrain models with synthetic or licensed data
  • Patch vulnerabilities
  • Document all changes affecting safety or compliance

Significant updates will be reflected on this page.

11. Responsible Use Reminder (For Users)

By using Spayo, you agree to:

  • Upload only images you have the rights to use
  • Avoid harmful or prohibited content
  • Respect third-party intellectual property
  • Not misuse generated output in illegal, fraudulent, or deceptive ways
  • Not attempt to reverse engineer our system

12. Contact Us

For questions about AI safety, content risks, or compliance:

📧 [email protected]
📧 [email protected]

We are committed to responsible innovation and a safe AI experience for all users.