How to moderate voice chats for F2P Multiplayer games with a small team

Social multiplayer games, especially those in VR, thrive on voice chat. It helps build relationships among players and increases the game’s immersion. However, it also opens the door to toxic behavior and harassment, which can be a leading cause of churn and low player retention.

As much as voice chat helps improve immersion, 30% of players said that they quit matches early to escape harassment, while 40% disabled voice chat altogether to avoid abuse. Epic Games was fined $520 million in 2022 and forced to turn off voice chat by default for kids.

Studio execs know that ensuring a safe, respectful community isn’t just an act of altruism, but a necessity for player loyalty and brand reputation. That being said, let’s explore what options you have when it comes to keeping your free-to-play multiplayer communities safe.

Community moderators

The immediately obvious choice for keeping your community safe it to have a moderation/safety team. These are staff members or volunteers who monitor for toxic behavior and moderate game communities. Historically, this has served as a deterrent to would-be toxic players, as they know their actions have consequences. Many successful studios run volunteer moderator programs, recruiting trusted, veteran players to help moderate community spaces.

Define Roles & Tools: if you empower mods, be clear about their responsibilities and the tools they need to manage the community. In-game, you might grant them the ability to “kick” players from sessions if a player is being abusive. Many social games (VRChat, Rec Room, Roblox, etc.) allow user-created rooms to have their own moderators, effectively letting the community police itself.

As your F2P social multiplayer game grows, community moderation gets a lot more complex. The sheer volume of players makes it almost impossible for moderators to catch numerous cases of toxicity. Unless you can secure more volunteers, the costs of hiring moderators rapidly add up as your multiplayer games grow in popularity.

AI-Powered Voice Moderation

If you have a small team with a free-to-play multiplayer game that has a growing community, AI-powered voice moderation might be worth considering. As mentioned in the above community moderators section, manual community moderation does not scale well.

Since the advent of LLMs and advancements in Artificial Intelligence, AI models have become exceptionally good at detecting toxicity. From OpenAI’s moderation API to more advanced moderation tools like PlaySafe and ToxMod, it has gotten significantly easier to detect toxicity in video games automatically. The primary difference between the two tools is that PlaySafe is a suite of community management tools built by game devs (in-game polling, AI NPCs, moderation). Consequently, ToxMod is solely a set of moderation tools. The beauty of AI moderation is that it can immediately take action against toxic players (e.g., banning a toxic player within seconds of them being toxic). This fosters a safer community, especially in F2P games…

Here’s how you can leverage AI-powered voice moderation:

Leverage real-time voice AI

Modern AI moderation tools can listen to voice chats in real time and flag, filter, or action toxic language, freeing you to do what you do best: making great games. Such AI tools use

Examples:

1. **Rec Room The social VR game Rec Room deployed real-time voice moderation using Modulate’s ToxMod across all public rooms and lobbies. As a result, instances of toxic voice chat fell by ~70%. This allowed a small safety team to effectively moderate a large multiplayer player base without hiring an army of moderators.**
2. **Roblox built an in-house voice moderation Machine Learning (ML) model called Roblox Sentinel and open-sourced it to benefit the industry.**
3. **BigBallers, Animal Rivals & DigiGods - All top of the meta quest store VR games use PlaySafe to moderate and keep their communities safe and gather feedback from players. As a result, games like Animal Rivals achieved a 92% reduction in toxicity. They grew to the top of the leaderboards over a few months by ensuring the rapidly growing community was friendly and approachable to new players.**

Automated enforcement & scalable policies

To get the AI to action on your behalf, you need to define what policies you have for community safety. This could be a list of phrases or words that you do not encourage in your community, or it could be a list of tools like PlaySafe that have robust policy creation systems that allow you to define the rules that govern your multiplayer game’s community.

User empowerment tools (mute, block & beyond)

A small internal team can’t be everywhere at once. Allowing players to mute or block players they either find toxic or intolerable empowers them to take control of their safety: many gamers, especially minorities and women (who often get targeted), pre-emptively mute strangers. Reports, mutes and blogs could also work hand in hand with AI moderation tools like PlaySafe  to calculate some kind of respect score (a measure of overall toxicity as well as player sentiment).

Design your game so that muting someone is seen as a standard safety precaution rather than an overreaction. To take full advantage of this, ensure players know how to mute themselves and others on voice chat. You can also consider muting audio by default for under-18 players (something Epic was forced to do in their FTC settlement).

Safety features in VR

For VR titles in particular, harassment can go beyond verbal—sometimes it’s physical. For example, invading personal space or unsolicited contact with your avatar in social VR games.

To curb this, you could implement features like personal space bubbles (which prevent others from “touching” a player’s avatar or getting too close if it makes sense for your game. Make such features on by default for minors. VRChat, for example, encourages users to use the built-in safety settings —muting, blocking, and a personal space bubble — to ward off trolls.

Educating your community (and the parents of minors) about these tools can also help keep children safe in online multiplayer social games. While AI goes a long way toward empowering users to protect themselves, it can set a solid foundation for safety in social games.

Lastly, you could also implement custom filters and voice masking. Custom filters allow players to alter how their voices sound in the game, which can make minorities and female players feel safer in social multiplayer games. While these features are advanced and not yet widely available, products like PlaySafe are constantly iterating to make them readily available, especially for F2P social multiplayer games.

The more harassment that players can effortlessly deflect on their own, the more manageable your community becomes with a small team.

Balancing safety, retention, and Brand reputation

A primary strategic concern for studio leaders is ensuring that any moderation effort does not kill the game’s fun. There is a delicate balance between creating a safe environment and over-policing players to the point where they feel like they are walking on eggshells.

Minor profanity should carry a lighter punishment (e.g., a warning) than major harassment/profanity, which should carry harsher punishments (e.g., a voice ban). Additionally, repeat offenders should have harsher punishments than first-time offenders.

Community health = Long-Term Revenue

Toxicity has a direct line to revenue and brand risk. Players who feel unsafe or unwelcome in your multiplayer game will churn, and they definitely won’t spend money on/in your game. Conversely, having a welcoming community drives growth, as players are more likely to recommend your game to their friends and play longer and spend more in it. In gaming, consider that if the voice chat is well moderated, players will use it more rather than disabling it, which, once again, boosts immersion and community engagement.

Conclusion

In conclusion:

• Prioritise player safety and well-being to foster a thriving, loyal community.
• Leverage AI-powered voice moderation for scalable and effective toxicity detection and enforcement.
• Empower players with user-friendly safety tools like mute, block, and personal space bubbles.
• Implement clear, nuanced moderation policies that differentiate between minor profanity and severe harassment.
• Understand that a healthy community directly translates to long-term player retention and increased revenue.