55% of UK children who game online play with people they have never met offline at least sometimes, and one in five do so all or most of the time they play, yet only 43% of parents set rules about who their child can play with (Ofcom, Growing Up Gaming 2026). That contact is often hostile: 41% of U.S. teen players have been called an offensive name while playing (Pew Research Center, Teens and Video Games Today 2024). Three-quarters of U.S. gamers aged 10 to 17 reported harassment in online multiplayer games in 2023 (ADL, Hate Is No Game 2023), a reminder that most of the risk travels through chat, the same channel our in-game proximity voice chat usage statistics track from the player side. We aggregated data from Ofcom, Pew Research Center, Thorn, the U.S. Federal Trade Commission (FTC), Australia’s eSafety Commissioner, ADL, NCMEC, Roblox, and national surveys from Ireland, Singapore and Japan.
TL;DR
- 55% of UK children who game online play with people never met offline at least sometimes, while only 43% of parents set rules on who they play with (Ofcom, Growing Up Gaming 2026).
- 41% of U.S. teen players have been called an offensive name while gaming, 48% of boys and 32% of girls (Pew Research Center 2024).
- Three-quarters of U.S. gamers aged 10 to 17 experienced harassment in online multiplayer games in 2023, up from 67% in 2022 (ADL 2024).
- 42% of Japanese minors have received a chat from a stranger while playing an online game (ChildFund Japan 2025).
- 36% of Singapore youths aged 10 to 18 played games with strangers sometimes or frequently (Singapore MCI 2024).
- 12% of Discord users and 8% of Roblox users aged 9 to 17 reported an online sexual interaction on that platform (Thorn, Youth Perspectives on Online Safety 2025).
- 87% of U.S. minors used online tools such as blocking or reporting after an online sexual interaction (Thorn 2025).
- Four in five parents of kids aged 5 to 9 use parental controls on some or all gaming devices (ESA, Holiday Guide 2025).
- The FTC alleged Fortnite voice chat was switched on by default in October 2017, against internal advice to make it opt-in (FTC, Epic Games complaint 2022).
- Three U.S. states settled with Roblox for a combined $35.8 million in April 2026 (Claims Journal 2026).
- 45% of Roblox’s 144 million daily active users had completed an age check by January 31, 2026 (Roblox, February 2026).
1. Who Is Playing: Children on Chat-Enabled Games
Gaming is close to universal for children, and most of it is social. 85% of U.S. teens aged 13 to 17 play video games, and 47% of teen players have made a friend online through a game they both play (Pew Research Center, Teens and Video Games Today 2024). A friendship that starts in a lobby is the same mechanism that lets a stranger start a conversation, so the social pull of games and the safety risk are two sides of one feature.
The platforms with the biggest child audiences are also the ones with built-in chat, which is why voice and text features, not game content alone, now sit at the center of the safety debate. A platform where 35% of age-checked users are under 13 (see section 7) cannot treat open voice chat as a neutral default. Content ratings tell the same story from another angle: most rated games carry a rating for everyone, older children or teens, so adults and minors can easily end up on the same servers.
| Metric | Value | Source |
|---|---|---|
| UK children aged 3 to 17 who play games on any device | 88% | Ofcom, Growing Up Gaming 2026 |
| U.S. teens aged 13 to 17 who play video games | 85% | Pew Research Center, Teens and Video Games Today 2024 |
| U.S. teen players who have made a friend online through a game they both play | 47% | Pew Research Center, Teens and Video Games Today 2024 |
| U.S. minors aged 9 to 17 who have ever used Roblox (share who use it at least daily) | 81% (41% daily) | Thorn, Youth Perspectives on Online Safety 2025 |
| Games rated by the ESRB in 2024 that received an E, E10+ or T rating | 89% | ESA, Holiday Guide 2025 |
Context: Pew’s teen survey was fielded from September 26 to October 23, 2023. Most recent available data: Pew Research Center, 2024.
2. Contact With Strangers in Game Lobbies
The headline figure is not a fringe behaviour. More than half (55%) of UK children who game online play with people they have never met offline at least sometimes, and 20% do so all or most of the time (Ofcom, Growing Up Gaming 2026). In a matchmade game, a “stranger” is the normal teammate: the system pairs players by skill and region, not by whether they know each other. That is why voice chat defaults carry so much weight.
National surveys from other countries land in a similar range, even with different age groups and question wording. In Japan, 42% of minors have received a chat from a stranger while playing an online game (ChildFund Japan, Research on Grooming via Online Games 2025). In Singapore, 14% of youths had engaged strangers outside of gaming, which is the step that moves a conversation off a moderated platform and out of reach of in-game reporting tools.
| Metric | Value | Source |
|---|---|---|
| UK children who game online and play with people never met offline, at least sometimes (all or most of the time) | 55% (20%) | Ofcom, Growing Up Gaming 2026 |
| Japanese minors who have received a chat from a stranger while playing an online game | 42% | ChildFund Japan, Research on Grooming via Online Games 2025 |
| Singapore youths aged 10 to 18 who played games with strangers sometimes or frequently | 36% | Singapore MCI, Survey on Online Gaming Activities 2024 |
| Singapore youths aged 10 to 18 who engaged strangers outside of gaming | 14% | Singapore MCI, Survey on Online Gaming Activities 2024 |
| Irish children aged 8 to 12 who gamed with strangers online | 38% | CyberSafeKids, Left To Their Own Devices 2024 |
Context: the Singapore survey was fielded door to door from October 2022 to February 2023 among 810 youths and their parents, covering youths who play at least once a month. Most recent available data: Singapore Ministry of Communications and Information (MCI), 2024.
3. Harassment and Abuse in Game Chat
Harassment is the most common harm, and it is unevenly spread. 41% of U.S. teen players say they have been called an offensive name while playing, rising to 48% of boys against 32% of girls (Pew Research Center 2024). More boys play, and more of them play with other people, so their exposure is larger. Girls are more likely to call harassment a major problem (33% vs 25% of boys), which suggests the harm is felt differently, not that it is smaller.
The trend line is not improving. ADL found that three-quarters of U.S. gamers aged 10 to 17 experienced harassment in online multiplayer games in 2023, up from 67% in 2022, and identity-based harassment, meaning abuse that targets who a player is rather than how they play, rose from 29% to 37% (ADL, Hate Is No Game: Hate and Harassment in Online Games 2023). ADL named League of Legends, PUBG and Call of Duty as the games where young people were most harassed, and its recommendations to studios specifically asked for stronger moderation tools for in-game voice chat.
| Metric | Value | Source |
|---|---|---|
| U.S. teen players who have been called an offensive name while playing (boys vs girls) | 41% (48% vs 32%) | Pew Research Center, Teens and Video Games Today 2024 |
| U.S. teen players who have been physically threatened (sent unwanted sexually explicit content) while playing | 12% (8%) | Pew Research Center, Teens and Video Games Today 2024 |
| U.S. teens who say harassment over video games is a problem for people their age (share calling it a major problem) | 80% (29%) | Pew Research Center, Teens and Video Games Today 2024 |
| U.S. gamers aged 10 to 17 who experienced harassment in online multiplayer games, 2023 vs 2022 | three-quarters vs 67% | ADL, Hate Is No Game 2023 (published February 2024) |
| U.S. gamers aged 10 to 17 who experienced identity-based harassment, 2023 vs 2022 | 37% vs 29% | ADL, Hate Is No Game 2023 (published February 2024) |
| Singapore youths aged 13 to 18 who experienced in-game bullying | 17% | Singapore MCI, Survey on Online Gaming Activities 2024 |
Outlier note: in Singapore, 48% of youths who were bullied in a game took no action and only 8% spoke to their parents (Singapore MCI 2024). Silence after an incident is a recurring pattern, and it limits how much of the problem shows up in platform reports. The ADL survey covered 1,971 Americans aged 10 to 45 and was fielded with Newzoo from August 4 to 17, 2023.
4. Sexual Exploitation Risk on Gaming Platforms
This is the rarest harm and the most serious one, so the denominators matter. Across all platforms, 1 in 4 U.S. minors aged 9 to 17 reported an online sexual interaction, including 1 in 6 preteens aged 9 to 12 (Thorn, Youth Perspectives on Online Safety 2025). Measured per platform, gaming services sit near or below the 11% average: 8% of Roblox users and 6% of Fortnite users reported such an interaction on that platform, while Discord, the voice and text app many players run alongside games, reached 12%. Social and messaging apps such as Snapchat and Kik reached 19%.
The lower per-user rate on game platforms does not make them low risk. Because so many children use Roblox, 7% of all minors in the survey had an online sexual interaction there, the highest share among the game-related platforms the survey lists. Report volumes show the same scale problem: Roblox filed 24,522 reports to NCMEC in 2024 (Roblox, response to the Louisiana Attorney General lawsuit, August 2025). Across every service, NCMEC received 20.5 million CyberTipline reports in 2024, down from 36.2 million in 2023, and nearly 100 reports of financial sextortion per day (NCMEC, 2024 in Numbers). A falling report total measures what platforms send, not how much abuse happens, so it should not be read as a safety improvement on its own.
| Metric | Value | Source |
|---|---|---|
| U.S. minors aged 9 to 17 who reported an online sexual interaction on any platform (aged 9 to 12; aged 13 to 17) | 1 in 4 (1 in 6; 1 in 3) | Thorn, Youth Perspectives on Online Safety 2025 |
| U.S. minors who had an online sexual interaction with someone they believed was an adult, any platform | 22% | Thorn, Youth Perspectives on Online Safety 2025 |
| Minor users of each platform who had an online sexual interaction there: Discord, Roblox, Fortnite (average across all platforms studied) | 12%, 8%, 6% (11%) | Thorn, Youth Perspectives on Online Safety 2025 |
| Japanese minors who experienced grooming, overall (boys vs girls) | 13% (18% vs 6%) | ChildFund Japan, Research on Grooming via Online Games 2025 |
| Reports Roblox submitted to NCMEC in 2024 (share of all reports submitted to NCMEC) | 24,522 (0.12% of 20.3 million) | Roblox, statement of August 2025 |
| Total CyberTipline reports received by NCMEC, 2024 vs 2023 | 20.5 million vs 36.2 million | NCMEC, 2024 in Numbers |
| Financial sextortion reports received by NCMEC per day, 2024 | nearly 100 | NCMEC, 2024 in Numbers |
Context: grooming means an adult building trust with a child in order to exploit them. In the ChildFund Japan survey, boys were three times as likely as girls to report it, which runs against the common assumption that girls are the main targets in gaming spaces. Thorn’s survey of 1,003 U.S. minors was fielded from November 12 to December 1, 2025.
5. Parental Controls and Family Rules
Parents name the right risk but act on a different one. 64% of UK parents of children who game online are concerned about their child talking to strangers through games or their chat functions, yet only 43% set rules about who their child can play with, the least common of the six gaming rules in the survey (Ofcom, Growing Up Gaming 2026). Time limits are easier to set and enforce than the social graph of a matchmade game.
Device-level controls are widespread among parents of young children: four in five parents of kids aged 5 to 9 use parental controls on some or all of the devices their kids game on (ESA, Holiday Guide 2025). Older children and other countries look different. In Ireland, 51% of children aged 8 to 12 have no parental restrictions on online contact such as chatting and gaming with people they do not know (CyberSafeKids, reported by The Irish Times, February 2026). For a broader view of tool adoption, see our parental controls in gaming statistics.
| Metric | Value | Source |
|---|---|---|
| UK parents of children who game online concerned about strangers via games or chat (concerned about bullying by other players) | 64% (60%) | Ofcom, Growing Up Gaming 2026 |
| UK parents who set rules on who their child can play with online vs parents who set time limits (parents who set no gaming rules at all) | 43% vs 56% (9%) | Ofcom, Growing Up Gaming 2026 |
| Parents of kids aged 5 to 9 who use parental controls on some or all gaming devices | 80% | ESA, Holiday Guide 2025 |
| Irish children aged 8 to 12 with no parental restrictions on online contact | 51% | CyberSafeKids, survey reported February 2026 |
| Singapore parents fully aware vs not at all aware of who their child games with | 31% vs 25% | Singapore MCI, Survey on Online Gaming Activities 2024 |
| Singapore parents concerned about their child’s gaming who used parental control tools | 22% | Singapore MCI, Survey on Online Gaming Activities 2024 |
| Japanese minors whose family has rules for online gaming | 50% | ChildFund Japan, Research on Grooming via Online Games 2025 |
Context: the Irish survey also found that only 39% of children aged 8 to 12 say their parents “check, monitor or control” what they do online, and that fewer than half of parents use parental controls over online activity (CyberSafeKids 2026).
6. Regulatory Enforcement: From the FTC Epic Case to Roblox Settlements
The FTC case against Epic Games was the first major enforcement action built partly around a voice chat default, and the complaint reads like a timeline of ignored warnings. According to the FTC, Epic switched voice chat on by default in Fortnite Battle Royale in October 2017, two months after its then UX director urged the company to avoid voice chat or make it opt-in “at the very least” (FTC, Epic Games complaint 2022). The FTC alleged that the UX team repeated the recommendation in June 2018, that parental PIN controls arrived only in June 2019, nearly two years after launch, and that a May 2018 internal email counted 834 support cases in the prior year containing “kill myself” and 485 containing “suicide”. Epic settled without admitting or denying the allegations.
What made the case a template was the remedy, not the money. The order requires voice and text communications to be off by default for children and teens, switched on only with a parent’s affirmative consent for under-13s, or the teen’s or a parent’s consent for teens (FTC press release, December 2022). The 2026 state settlements with Roblox follow the same logic: West Virginia’s deal requires that all chat be blocked until age verification is complete and that adults reach users under 16 only through verified trusted friends (West Virginia Attorney General, April 2026). Our children’s privacy enforcement statistics cover the wider COPPA and data-protection penalty record.
| Metric | Value | Source |
|---|---|---|
| Month Epic made voice chat on by default in Fortnite Battle Royale, per the FTC complaint | October 2017 | FTC, Epic Games complaint 2022 |
| Fortnite support cases in the year before May 2018 containing the phrase ‘kill myself’ vs the word ‘suicide’ | 834 vs 485 | FTC, Epic Games complaint 2022 |
| U.S. children aged 10 to 12 vs teens aged 13 to 17 who played Fortnite weekly, in a 2019 survey cited by the FTC | 53% vs 33% | FTC, Epic Games complaint 2022 |
| Combined Roblox child-safety settlements with Alabama, Nevada and West Virginia, April 2026 | $35.8 million ($12.2M, $12.5M, $11.1M) | Claims Journal, April 22, 2026 |
| West Virginia settlement money for a state-based internet safety specialist | $2.4 million over six years | West Virginia Attorney General 2026 |
| U.S. states with ongoing child-safety lawsuits against Roblox listed at the time of the April 2026 settlements | 5 (Tennessee, Texas, Iowa, Florida, Nebraska) | Claims Journal, April 22, 2026 |
| Maximum civil penalty eSafety can seek for breaching Australia’s online safety codes and standards | AUD 49.5 million | eSafety Commissioner, September 2025 |
| Maximum fine under the UK Online Safety Act | 18 million pounds or 10% of qualifying worldwide revenue, whichever is greater | UK Government, Online Safety Act explainer |
Context: the West Virginia settlement totals $11,080,000 and was reached before the state filed a lawsuit. Of that, $500,000 funds safety workshops and $1.5 million a three-year public safety campaign (Claims Journal, April 2026).
7. Platform Safety Measures: Age Checks and Chat Defaults
The industry response has shifted from optional settings to gated access. Roblox made an age check mandatory for chat worldwide on January 7, 2026, and 45% of its 144 million daily active users had completed one by January 31, 2026 (Roblox, Moving Beyond Self-Reported Age 2026). Users under 9 need parental consent to chat, and age bands limit who can talk to whom, with the stated aim of keeping children under 16 from communicating with adults (Roblox, Age Checks Required to Chat, January 2026).
Regulators are now testing these claims rather than accepting them. Under commitments made to Australia’s eSafety Commissioner in September 2025, voice chat is not allowed between adults and 13 to 15 year olds, under-13s were already barred from voice chat, and accounts for under-16s are private by default (eSafety Commissioner, September 2025). Children’s own behaviour shows why platform tools matter: 87% of U.S. minors turned to online tools such as blocking and reporting after an online sexual interaction, nearly twice the share who sought offline support (Thorn 2025).
| Metric | Value | Source |
|---|---|---|
| Roblox daily active users who had completed an age check (facial age estimation or ID) by January 31, 2026 | 45% | Roblox, Moving Beyond Self-Reported Age 2026 |
| Age-checked Roblox daily active users aged under 13, 13 to 17, and 18 or older | 35%, 38%, 27% | Roblox, Moving Beyond Self-Reported Age 2026 |
| Daily active users in Australia, New Zealand and the Netherlands who completed an age check after the December 2025 early rollout | more than 50% | Roblox, Age Checks Required to Chat 2026 |
| Mean Absolute Error (average gap between estimated and real age) of Roblox’s facial age estimation for users under 18 | 1.4 years (company-reported result of UK Age Check Certification Scheme testing) | Roblox, Age Checks Required to Chat 2026 |
Context: all Roblox figures in this section are the company’s own, self-reported and not independently audited. The 1.4-year error figure is an average, so individual estimates can miss by more, which matters most near the age-band boundaries at 9, 13 and 16.
Summary: Child Safety in Online Game Voice Chat by the Numbers
| Metric | Value | Source |
|---|---|---|
| UK children who game online and play with people never met offline, at least sometimes | 55% | Ofcom, Growing Up Gaming 2026 |
| UK parents who set rules on who their child can play with vs time limits | 43% vs 56% | Ofcom, Growing Up Gaming 2026 |
| U.S. teen players called an offensive name while playing | 41% | Pew Research Center 2024 |
| U.S. gamers aged 10 to 17 harassed in online multiplayer games, 2023 vs 2022 | three-quarters vs 67% | ADL 2024 |
| U.S. gamers aged 10 to 17 who experienced identity-based harassment, 2023 | 37% | ADL 2024 |
| Japanese minors who received a chat from a stranger in an online game | 42% | ChildFund Japan 2025 |
| Singapore youths aged 10 to 18 who played with strangers sometimes or frequently | 36% | Singapore MCI 2024 |
| Irish children aged 8 to 12 who gamed with strangers online | 38% | CyberSafeKids 2024 |
| U.S. minors reporting an online sexual interaction on any platform | 1 in 4 | Thorn 2025 |
| Minor users with an online sexual interaction on Discord, Roblox, Fortnite | 12%, 8%, 6% | Thorn 2025 |
| Total NCMEC CyberTipline reports, 2024 | 20.5 million | NCMEC 2025 |
| Parents of kids aged 5 to 9 using parental controls on gaming devices | 80% | ESA 2025 |
| Irish children aged 8 to 12 with no parental restrictions on online contact | 51% | CyberSafeKids 2026 |
| Fortnite voice chat switched on by default, per the FTC | October 2017 | FTC 2022 |
| Combined Roblox state settlements, April 2026 | $35.8 million | Claims Journal 2026 |
| Maximum civil penalty under Australia’s online safety codes | AUD 49.5 million | eSafety Commissioner 2025 |
| Roblox daily active users who completed an age check by January 31, 2026 | 45% | Roblox 2026 |
Methodology and Sources
- Ofcom, Growing Up Gaming research (October 13, 2026): Ofcom’s site blocked automated access, so figures were read in coverage by Advanced Television and LBC, which report a sample of 1,719 children and 2,522 parents of 3 to 17 year olds.
- Pew Research Center, Teens and Video Games Today (May 9, 2024)
- ADL, Hate Is No Game: Hate and Harassment in Online Games 2023 (press release, February 2024)
- Thorn, Youth Perspectives on Online Safety, 2025 (published July 2026)
- NCMEC, 2024 in Numbers (2025)
- FTC, complaint against Epic Games (December 2022) and FTC press release on the Epic Games settlement (December 2022)
- West Virginia Attorney General, Roblox settlement (April 21, 2026) and Claims Journal, Roblox settles with states for $35.8 million (April 22, 2026)
- eSafety Commissioner, Roblox commitments under Australia’s online safety codes (September 2025)
- UK Government, Online Safety Act explainer
- Roblox newsroom: Moving Beyond Self-Reported Age (February 5, 2026), Age Checks Required to Chat (January 2026), response to the Louisiana Attorney General lawsuit (August 2025)
- ESA, Holiday Guide 2025
- ChildFund Japan, Research on Grooming via Online Games (2025): online survey of 1,616 people aged 13 to 25, including 1,049 minors, fielded from November 28 to December 5, 2024.
- Singapore Ministry of Communications and Information, Survey on Child’s Online Gaming Activities (February 13, 2024)
- CyberSafeKids (Ireland): Left To Their Own Devices Trends and Usage Report, reported by TheJournal.ie (September 3, 2024); 2026 survey of more than 3,300 children aged 8 to 12, reported by The Irish Times (February 10, 2026)
- Data watch: NCMEC gives 20.5 million total CyberTipline reports for 2024, while Roblox puts total reports submitted to NCMEC at 20.3 million; the gap likely reflects reports from the public versus reports from online platforms, so the two totals should not be mixed. CyberSafeKids’ 2024 figure for children aged 8 to 12 contacted by a stranger in an online game was reported as 61% in some outlets and “almost 65%” in others, so we left it out of the tables. The Claims Journal list of five states still suing Roblox was a snapshot from April 2026 and may have changed since. The FTC Epic Games figures are allegations from a December 2022 complaint that Epic neither admitted nor denied, and the Fortnite age survey it cites dates from 2019. ESA’s 80% parental controls figure is published without a stated sample or field dates. All Roblox figures are self-reported. Pew’s teen survey (fieldwork 2023) and Singapore’s survey (fieldwork 2022 to 2023) are the most recent available data for their questions; a new Pew teen gaming wave would supersede the U.S. figures in section 3.
Last updated: November 6, 2026. We update this roundup quarterly, and the next refresh is expected when Ofcom publishes its 2026 Children and Parents: Media Use and Attitudes report and Roblox issues its next quarterly shareholder letter with updated age-check figures.