Trust

AI Disclosure

How we use AI to generate school-photo style portraits.

Last updated: January 1, 2026

SmilePlease uses artificial intelligence to transform your photos into professional school-style portraits. We believe you have the right to understand how this technology works and make informed decisions about using it.

The AI we use

Google Gemini

SmilePlease uses Google's Gemini image generation models to create portrait variations. Gemini is developed by Google DeepMind and is designed with safety and quality guardrails.

We chose Gemini because it produces high-quality, natural-looking portraits while maintaining strong safety controls for images involving children.

How it works

  1. Upload: You provide a photo of your child (or yourself).
  2. Processing: Your photo is sent securely to Google's Gemini AI service along with our instructions for creating school-portrait styling.
  3. Generation: Gemini generates 6 unique variations, each with slightly different backgrounds, lighting, or subtle pose adjustments.
  4. Review: You see watermarked previews and choose which portraits to purchase.
  5. Delivery: Selected portraits are printed and shipped (or delivered digitally).

The entire generation process typically takes about 60 seconds.

What AI can and cannot do

What our AI does well

  • Creates professional-looking school portrait styling
  • Generates multiple variations so you can find the perfect one
  • Maintains your child's likeness and natural appearance
  • Produces high-resolution images suitable for printing

Known limitations

  • Visual imperfections: AI may occasionally produce visual artifacts (unusual textures, distortions, or imperfections).
  • Variation: Results are not perfectly predictable—each generation is unique.
  • Source quality: Results depend heavily on the quality of your uploaded photo (lighting, resolution, clarity).
  • Accessories: Glasses, hats, or unusual hairstyles may not be rendered consistently.
  • Complex backgrounds: Photos with busy backgrounds may produce less consistent results.

Recommendation: Always review all 6 variations carefully before purchasing. The watermarked previews show exactly what you'll receive.

Likeness commitment: SmilePlease is designed to change setting/composition while preserving your child's identity. If final delivered output materially fails to match approved preview/likeness expectations, we'll make it right under our Refund Policy.

Safety measures

We've implemented multiple layers of safety to ensure appropriate use:

  • Careful instructions: We give the AI specific instructions that request professional school-portrait styling, limiting the AI's creative latitude to appropriate outcomes.
  • Google's safety filters: Gemini includes built-in content safety measures that block inappropriate content generation.
  • Output review: Generated images pass through safety checks before being displayed.
  • Use restrictions: Our Acceptable Use Policy prohibits misuse including deepfakes, impersonation, and harmful content.

Human oversight

While AI generates the portraits, humans remain in control:

  • You choose: You review all variations and decide which ones to purchase. Nothing is printed or charged without your approval.
  • Quality review: Our team monitors system outputs for quality and safety issues.
  • Abuse reporting: We review all reports of misuse and take action when necessary.
  • Instruction design: Our instructions are designed and refined by humans to produce appropriate results.

Bias and fairness

We're committed to ensuring our AI works well for everyone:

  • Diverse testing: We test portrait generation across a wide range of skin tones, hair types, facial features, and ages.
  • Continuous improvement: We monitor results and refine our prompts to improve quality for all users.
  • Feedback welcome: If you experience quality issues that may be related to bias, please contact us. Your feedback helps us improve.

Note: All AI systems can reflect biases from their training data. While Google works to minimize bias in Gemini, no AI system is perfect. We take this seriously and are committed to addressing issues as they arise.

Your data and training

We do NOT use your photos to train AI models.

Your uploaded photos and generated portraits are used only to provide you with the SmilePlease service:

  • Photos are sent to Gemini solely for generating your portrait variations.
  • According to Google's data usage policies for Gemini AI service, user-submitted content is not used to train or improve Google's models.
  • Uploaded photos are deleted within 10 days of generation. Generated portraits are deleted after 30 days (or 12 months if you purchase and opt in to Memory Keeper).
  • See our Privacy Policy for complete details on data handling.

Digital fingerprint for AI images (C2PA)

All portraits generated by SmilePlease are AI-generated images, not traditional photographs. We add an invisible digital label to show they were created by AI, using the C2PA (Content Authenticity Initiative) standard:

  • Content Credentials: Purchased prints and digital downloads include an invisible digital label indicating they were AI-generated by SmilePlease using Google Gemini.
  • Verifiable provenance: This metadata can be verified using tools like Content Credentials Verify or compatible browser extensions.
  • Transparency: Portraits should not be represented as authentic photographs in contexts where authenticity matters (journalism, legal proceedings, etc.).
  • Our Acceptable Use Policy prohibits deceptive use of generated portraits.

Third-party AI

When you use SmilePlease, your photo is processed by Google's Gemini service. This means:

  • Your photo is transmitted to Google's servers for AI processing.
  • Google's Privacy Policy and Terms of Service apply to this processing.
  • We use Gemini's API, which is designed for business use with appropriate privacy protections.

Questions about AI

Have questions about how we use AI? We're happy to explain. Contact us at Contact us.

Updates to this disclosure

We will update this disclosure if we change AI providers, significantly modify our prompts, or make other material changes to how AI is used in our service. Check the "Last updated" date above for the most recent version.

Related policies