Skip to main content

Archived Article — The Daily Perspective is no longer active. This article was published on 7 March 2026 and is preserved as part of the archive. Read the farewell | Browse archive

Gaming

Roblox deploys AI to rephrase profanity rather than censor chat outright

New real-time language filter aims to keep conversations readable while maintaining safety standards on the gaming platform

Roblox deploys AI to rephrase profanity rather than censor chat outright
Image: Engadget
Key Points 3 min read
  • Roblox introduced real-time AI chat rephrasing that converts profanity into more respectful language rather than censoring it with #### symbols
  • The feature is limited to age-checked users in similar age groups; users still face penalties for repeated profanity violations
  • Roblox also enhanced its text filters to better detect attempts to evade moderation, reducing false negatives by 20 times for personal information exchanges
  • The move follows the company's mandatory global age verification system launched in January 2026 amid dozens of lawsuits alleging inadequate child protection

Roblox has introduced a real-time, AI-powered chat rephrasing feature that automatically replaces banned words with more respectful language, rather than displaying hash marks. The shift represents a pragmatic acknowledgment thatstrings of "####" can disrupt conversations and make them hard to follow, particularly during active gameplay when quick communication matters.

A message reading "Hurry TF up!" would previously have appeared as "####" but will now be rephrased to "Hurry up!"Everyone in the chat is notified when a message has been rephrased, and the sender sees what language was filtered. The feature currently targets profanity and is rolling out across all languages supported by Roblox's automatic translation system.

However, the rephrasing tool operates within strict constraints.Users who continue to violate the platform's profanity policy face the same consequences, even when their messages are rephrased.The company's multilayered safety system remains in effect for more serious behaviour. The feature isavailable exclusively for in-experience chat between age-checked users in similar age groups and their Trusted Connections.

Roblox's Chief Safety Officer described the approach as balancing user experience with safety norms. "This system acts as a steering wheel rather than a stop sign; if a message violates policy, everyone in chat is notified that text has been rephrased to keep things civil while helping keep chat respectful without losing authentic feel or flow."

The company has also enhanced its text-filtering infrastructure to combat evasion tactics.Early results show the system is getting better at detecting leetspeak and sophisticated attempts to bypass filters, allowing the company to reduce false negatives when sharing or soliciting personal information by 20 times.

The announcement arrives amid intensifying regulatory pressure and legal scrutiny.The rephrasing feature follows a string of child-safety lawsuits from state attorneys general, including Texas, Kentucky, and Louisiana, and comes after newly mandated facial verification for chat access. Several of these lawsuits have alleged that the platform failed to prevent predatory behaviour and grooming.The platform has 151.5 million daily active users, many of them minors, and several U.S. states have filed lawsuits accusing it of exposing children to risks such as grooming and inappropriate content.

The announcement comes shortly after Roblox introduced mandatory facial verification for access to chats, following a wave of lawsuits over child safety from attorneys general of Texas, Kentucky, and Louisiana, among others.Facial age verification became mandatory in Australia, New Zealand, and the Netherlands on December 1, 2025, and became mandatory globally on January 1, 2026.

Evaluating this development requires acknowledging genuine trade-offs. From a safety perspective,in experiments last year, Roblox saw that issuing in-experience text chat notifications and time-outs resulted in a 5 per cent reduction in filtered chat messages and a 6 per cent reduction in consequences from abuse reports, suggesting that real-time feedback and guidance can encourage compliance with platform norms. Conversely, critics argue that rephrasing tools risk softening enforcement; some worry that making violations less visible could reduce user awareness of boundaries.

Roblox acknowledges this is "a bold new approach" and that the company "won't always rephrase things perfectly," signalling an experimental mindset. The company consulted with its Teen Council during development, embedding younger users' voices into the design.

The central question for policymakers, parents, and the platform itself is whether technological intervention can substitute for institutional accountability. Rephrasing improves user experience and may nudge behaviour toward civility. Yetas long as there is a profit motive to keep engagement high and moderation costs low, the tension between safety and growth persists, and verification tools are helpful barriers but no substitute for active supervision. Roblox's multi-layered approach addressing age-gating, real-time moderation, and enhanced filter detection reflects reasonable thinking about harm reduction. Whether it goes far enough remains contested in ongoing litigation.

Sources (7)
Tanya Birch
Tanya Birch

Tanya Birch is an AI editorial persona created by The Daily Perspective. Reporting on organised crime, family violence, and court proceedings with meticulous legal precision. As an AI persona, articles are generated using artificial intelligence with editorial quality controls.