Riot Games is bringing a background launch of a voice evaluation system to VALORANT next month to help train language models for a beta later this year.
Voice comms in VALORANT are an essential part of the experience, allowing teams to work together to outsmart their opponents. Unfortunately, it is not uncommon for players to be toxic in voice chat, significantly ruining the experience for other players. But Riot is working on a new voice evaluation system that should improve the experience for all players.
In April, Riot updated its Privacy Notice and Terms of Service, allowing the recording and evaluation of in-game voice communications in VALORANT when a report for abusive voice comms was received. The organization also confirmed earlier this year that “voice evaluation would provide a way to collect clear evidence that could verify any violations of behavioral policies before we can take any action.” This change even allows Riot to share the reasoning for a penalty.
On July 13, Riot will begin a “background launch of the voice evaluation system” to help train language models and prepare for a beta launch later in 2022. The background launch will be in North America and English only. The voice evaluation will not be used for disruptive behavior reports during this time, which will only begin in the future beta.
The developers are aware that the system must be effective and accurate before a widespread launch and have “systems in place” to correct any “false positives.” Riot also confirmed this is new tech and there will likely be “growing pains,” but a safer and inclusive environment is worth the trouble.