AIGC Compliance Disclosure

Model disclosure and safety controls for Couple Goals AI
Jul 31, 2026

Model Disclosure

Our current couple photo generation service uses Google Gemini 3.1 Flash (Nano Banana 2) as the underlying image model.

Safety Controls

Before generation, prompts are checked by our automated content safety flow using Waffo Prompt Sift. This system scans for prohibited content across all six categories detailed in our AIGC Use Policy — NSFW content, violence, hate speech, CSAM, deepfakes, and copyright infringement. We maintain semantic analysis (enforce mode) to catch ambiguous or borderline prompts that keyword-only filters might miss.

We also maintain a public reporting channel and moderation workflow through our AIGC Use Policy.

Data Handling

We only share the minimum data required to generate your requested image, such as uploaded photos, scene selections, and prompt text. We do not use uploaded photos to train our own AI models.

Updates

If our underlying model changes, we will update this disclosure to reflect the current production model.

AIGC Compliance Disclosure