Hello VRChat Team,

I'd like to suggest a new feature called "Parent Mode" to enhance safety for younger users while preserving the freedom and fun of VRChat for everyone else. This mode would only activate if a parent enables it on a child's account, leaving all other accounts unchanged—like they are today. No mandatory age verification or global restrictions; the platform stays the same for normal users. The goal is to make VRChat more family-friendly for those who want it, potentially attracting more users, while addressing concerns like bad language, creepy interactions, and unsafe content. Below is a detailed outline of how it could work.

Core Functionality: Privacy-Respecting AI Recording

Recording activates only in mixed kid-adult interactions (e.g., when a child with Parent Mode enabled talks to an adult). Kid-kid or adult-adult chats remain unrecorded to maintain privacy and the platform's open nature.

All recordings are temporary and automatically deleted after 48 hours. If no report is filed, nothing is saved.

If a report is submitted (e.g., for bad language or inappropriate behavior), the relevant clip is retained for 30 days for review by moderators or developers. This ensures evidence is available for legitimate issues without long-term storage.

Reporting System with Anti-Abuse Measures

Anyone can report an interaction or content via a simple button. Reports trigger a review of the saved clip (if applicable).

To prevent spam or false reports:

The first 3-4 false reports from a user are treated as accidents (no penalty).

After that, the user's reporting ability is temporarily disabled: 2 weeks for the first offense, 3 weeks for the second, 4 weeks for the third, and increasing durations up to permanent if the behavior continues.

If a report is valid, developers review the playback, confirm the issue, and take action (e.g., blocking the offender for kids). If the AI made a mistake, it learns from the review to improve accuracy.

Avatar Filtering for Safety

In kid-adult interactions, the AI scans adult avatars in real-time. Only extreme cases (e.g., those with graphic violence, nudity, or highly disturbing elements) are blocked or replaced with a safe placeholder (e.g., a cartoon blob).

Not all "scary" avatars are blocked—context matters (e.g., a mild zombie mask is okay, but a chainsaw head is not). The system focuses on truly harmful content, not mild edginess.

If an avatar is blocked incorrectly, users can report it as a false positive. Developers review and adjust, ensuring the AI gets smarter over time without over-censoring.

World Filtering and Ratings

Introduce an optional "Kid-Friendly" tag for world creators. If tagged, the AI scans the world for inappropriate elements (e.g., swearing in scripts, gore props, or hidden adult content).

If the scan confirms it's safe (e.g., no extreme violence, nudity, or drugs), the world appears in kid searches. If not, the tag is automatically removed—no punishment for creators, just an accurate filter.

For untagged worlds, the AI auto-rates them based on content: All Ages (safe for everyone), 9+ (mild themes like cartoon guns or non-graphic drinking), 13+ (some swearing or light violence), or 17+ (adult themes like smoking, heavy blood, or romance).

Kids with Parent Mode enabled only see worlds matching their age rating. Attempting to join a higher-rated world (e.g., via direct link) results in a gentle block: "This world is for older users—ask a parent to review?" No access granted without override.

Parent Controls and Integration

Parents link a companion app (mobile or PC) to the child's account for easy toggling of Parent Mode on/off.

The feature only works if Parent Mode is enabled—without it, the account behaves exactly like a normal one, with full access to everything.

Alerts are minimal: Parents only receive notifications for confirmed issues (e.g., a kid gets blocked from an unsafe interaction). No constant monitoring or daily logs.

Customizable filters: Parents can adjust sensitivity (e.g., allow mild scary avatars or cartoon violence) while keeping core protections like anti-swearing active.

This feature balances safety with VRChat's creative freedom, encouraging more families to join without affecting regular users. It draws inspiration from successful systems in other platforms but is tailored to VRChat's unique, user-generated environment. Privacy is prioritized with short retention periods and no universal recording. Implementation could start small (e.g., beta testing with voice filters) and expand based on feedback.

Thank you for considering this! I'm a long-time user and believe it could make VRChat even better.