I want to start with something that happened in 2024 that didn’t get nearly enough attention.
A woman at a major game studio posted publicly about receiving sustained targeted harassment after shipping a title. Not criticism of the game. Not feedback, even harsh feedback. Coordinated harassment aimed at her personally, her identity, her appearance, her right to work in the industry. Her colleagues at several other studios reported similar experiences that same month. Most of their employers said nothing publicly.
That’s not a one-off. It’s a pattern that’s been documented, measured, and tracked for years. And despite the fact that multiple major publishers have deployed real moderation infrastructure, the AI tools covered in earlier articles in this series, the targeting of specific groups within gaming has gotten worse by some measures, not better.
This article is about who gets hit hardest, what the data actually shows, why it’s happening, and which games and studios are doing something real about it versus which ones are staying quiet and hoping nobody notices.
The Numbers First
We covered some of this data in Article 8, but the identity-based harassment statistics deserve more than a paragraph. The full picture is worse than most people realize, and the trend lines are going in the wrong direction.
Black gamers. The Anti-Defamation League has tracked gaming harassment annually since 2019. In 2020, 31% of Black adult gamers reported race-based harassment. By 2023, that number was 50%. That’s a 60% increase in three years, during a period when the industry was investing more in moderation than at any prior point in its history. The tools are helping overall harassment rates. They’re not keeping pace with the rise in targeted racial harassment.
Women. The ADL 2023 data shows 48% of women experienced gender-based harassment. That’s the broad population number. In competitive environments, it’s worse. A 2026 academic study examining competitive gaming communities found that approximately 87% of women and transgender and gender-diverse players reported gender-related harassment in competitive settings. The gap between “all gaming” and “competitive gaming” is significant, ranked modes, high-stakes lobbies, and visible public play all intensify targeting.
LGBTQ+ gamers. GLAAD’s 2024 Gaming Report is the most detailed dataset we have on LGBTQ+ gaming experiences. Key findings from 1,452 surveyed players:
- 52% of LGBTQ+ gamers experienced harassment, compared to 38% of non-LGBTQ+ players
- 42% actively avoided games they anticipated would be hostile to them
- 61% reported being uncomfortable using voice chat at all
- LGBTQ+ gamers were significantly more likely to report that harassment caused them to spend less time gaming overall
That 61% voice chat discomfort figure is striking. Voice chat is how a huge portion of multiplayer gaming communication works, team coordination, callouts, social connection. For the majority of LGBTQ+ players to have opted out of it entirely because the risk calculus doesn’t work in their favor isn’t a preference. It’s an adaptation to a hostile environment.
Jewish gamers. ADL data shows 34% of Jewish gamers faced identity-based harassment in 2023, up from 22% in 2021. That increase tracks with broader trends in online antisemitism that have been documented across platforms, gaming communities aren’t insulated from what’s happening in broader internet culture.
Teen players in marginalized groups. ADL’s youth survey found 75% of players aged 10-17 experienced some form of harassment, with identity-based targeting being common. Critically, only 39% of youth reports to game companies resulted in any meaningful action. For a teenager experiencing targeted harassment based on race or gender, filing a report and getting no response isn’t a neutral outcome. It’s a signal about whose experiences the industry takes seriously.
What Happens When People Get Targeted
The behavioral response data is where the cost of this becomes concrete.
From the GLAAD report: 42% of LGBTQ+ gamers avoided games they anticipated would be hostile. They didn’t just quit games after having bad experiences. They preemptively excluded games from consideration based on community reputation. The harassment is shaping purchase decisions before a single session is played.
The ADL’s username experiment documented something that sounds almost too crude to be real but has been replicated in various forms: accounts with identifiably female usernames, racially coded names, or LGBTQ+-associated identifiers receive measurably more harassment than neutral usernames playing the same games in the same modes. The targeting isn’t random. Players are scanning for identity signals and using them as selection criteria.
Riot’s internal data on League of Legends first-time player churn, 320% more likely to quit after a toxic early experience, is an aggregate number. The churn rate for players from targeted groups who experience identity-based harassment in their first sessions is almost certainly higher, though Riot hasn’t published a breakdown at that level of specificity.
What this adds up to: gaming is actively losing players from specific groups to toxicity at higher rates than the average. The people most likely to benefit from more diverse representation in gaming communities are the people most likely to be pushed out of those communities by the existing ones.
Gamergate 2.0 and What It Told Us
In 2024, a campaign to boycott games from a narrative consulting company called Sweet Baby Inc. spread across gaming platforms. The campaign escalated into targeted harassment of employees at studios with no connection to the company. Developers received death threats. Personal information was posted publicly. Several people left the industry.
The game community press called it Gamergate 2.0, a reference to the coordinated harassment campaign from 2014 that drove several women out of game development and remains a reference point for how organized toxicity operates in gaming spaces.
What was notable about the 2024 iteration wasn’t just that it happened. It was the response, or lack of one, from major publishers.
Most stayed silent. Several studios whose employees were being targeted declined to make public statements. The argument, presumably, was that engaging would amplify the situation. The counterargument, the one female developers and LGBTQ+ developers made publicly, was that silence from their employers in the face of coordinated targeting of their staff sent its own message about whose safety was a priority.
A few studios were exceptions. Riot Games had already built a public posture around community safety that gave them credibility to speak on this. Bungie put out a statement. Individual executives at a handful of other companies responded personally. But as a proportion of the major publishers affected, the silence was the dominant response.
The Black Myth: Wukong release in August 2024 generated another wave of targeted harassment around gender and racial diversity discussions in gaming media. The Dragon Age: The Veilguard backlash became another venue for coordinated hostility toward developers perceived as representing diversity in game development.
What 2024 demonstrated is that the moderation infrastructure improvements in games, ToxMod, GGWP, automated text systems, operate at the in-game session level. They don’t address what happens when coordinated harassment campaigns move between gaming platforms, Discord servers, Twitter threads, and Reddit communities. Behavior that gets flagged in one venue continues in another. The tools haven’t caught up to the multi-platform reality of how organized harassment works.
The Games That Are Actually Doing Better
I don’t want to leave this at “here’s how bad it is.” The data shows what’s working, too.
Final Fantasy XIV has the most consistently cited reputation for being welcoming to players from marginalized groups of any major MMORPG. The reasons are documented and specific, not just vibes:
The game’s lore and art direction have historically included queer relationships and diverse character design in ways that attracted LGBTQ+ players early in the game’s community development. Once that population is established and vocal, community norms shift. Veteran FFXIV players actively enforce a culture of inclusion that is self-perpetuating, not because Square Enix mandated it through moderation alone, but because the community itself treats harassment of other players as a community violation worth responding to.
The “Mentors” system formally recognizes veteran players who help new ones. The new player protection infrastructure (Sprout markers, mentor chat channels) creates positive first experiences for players who don’t yet know anyone in the game. First impressions in safe environments produce very different long-term retention outcomes than first impressions in hostile ones.
Deep Rock Galactic eliminates most of the structural conditions that produce targeting. No PvP means no performance-ranked conflict. Cooperative design means a struggling teammate is a shared problem, not a blame target. The game’s humor and tone, space dwarves, “Rock and Stone,” management satire, attracts players who genuinely don’t take themselves too seriously. Developer Ghost Ship Games maintains an active community presence that models the tone they want from the player base.
The DRG subreddit is legitimately one of the more pleasant gaming communities online. That’s not accidental. It’s the cumulative result of game design, developer community management, and a player base self-selected by the game’s tone and structure.
Warframe is worth mentioning because it’s free-to-play and has every structural reason to attract the chaotic player dynamics that free-to-play games often generate. The reason it hasn’t comes down to developer culture at Digital Extremes, consistent direct engagement with the player community, genuine responsiveness to concerns, and a progression system that keeps invested players around long enough to develop community attachment.
The veteran-to-new-player help culture in Warframe is something players from marginalized groups have noted and compared favorably to other games. When new players are met with offers of help rather than assessment of their performance metrics, the first-session experience is fundamentally different.
What Publishers Are Actually Doing, And Not Doing
Riot Games has been the most explicit and consistent of any major publisher in their public stance on community harassment.
Their statement on hardware bans for toxic players in May 2024, expanding a tool previously reserved for cheaters to players who engage in hate speech and targeted harassment, was accompanied by a direct statement from studio head Anna Donlon:
“If you need to make truly evil statements in the guise of regular shit talk to enjoy gaming, please play something else. We won’t miss you.”
That’s a public commitment from a named executive at a major studio. It’s significantly different from a press release about “improving community standards.” It names the behavior, names the consequence, and makes clear there’s no welcome mat.
Riot also updated their accountability policies in December 2024 to extend to content creators, streamers and community figures who model toxic behavior in front of their audiences face the same enforcement as players who exhibit it in-game. The recognition that influential community figures shape norms, not just experience them, is a meaningful policy development.
Activision’s CoD code of conduct update in 2024 added a prohibition on “amplifying or promoting discriminatory movements”, explicitly covering organized campaigns that use gaming spaces to spread coordinated hate. That’s a direct response to the Gamergate 2.0 style of harassment that operates through signals and dog whistles rather than explicit slurs that keyword filters catch.
Ubisoft’s Fair Play Program takes a different angle, educational modules designed with psychologists to help players recognize their own toxic triggers and behavior patterns. The argument is that some toxic behavior comes from players who’ve normalized it through exposure and don’t fully recognize what they’re doing. The research backing this is mixed, but the attempt to address root causes rather than only symptoms is worth acknowledging.
The publishers who said nothing in 2024: A significant portion of the major studios. I’m not going to name them individually here, because the list changes and “said nothing publicly” covers a range of behavior, from active internal work not communicated externally, to genuine institutional indifference. What I’ll say is that the absence of public statements from major publishers whose employees were targeted in 2024 was noticed and documented. It’s part of the transparency picture the TAG Community Safety Score will assess.
The Bystander Effect Nobody Talks About
The ADL research found that 88.1% of players witnessed toxicity happening to someone else, even among the fraction who didn’t experience it directly.
Gaming research on bystander behavior shows the same patterns as bystander research in other contexts: the more people who are present and don’t intervene, the less likely any individual is to act. In a lobby where someone is being targeted with slurs and six other players say nothing, the message to the target is that the silence is the consensus. It often isn’t, most players in those lobbies are simply defaulting to the inaction bias that bystander research predicts. But the target can’t distinguish between active indifference and passive discomfort.
Some games have tried to address this explicitly. Overwatch 2‘s commendation system, Rainbow Six Siege‘s reputation system, and League of Legends‘ Honor system all create positive feedback loops for prosocial behavior, giving bystanders a mechanism that isn’t just “say something and possibly invite targeting yourself.” A click that commends a teammate who handled harassment with grace is an action. It’s a small one. But behavioral research suggests that small positive actions reduce the psychological friction toward larger ones over time.
The “nudge” features that PlayStation and some other platforms have deployed, prompting players to reconsider before sending flagged messages, operate on similar logic. Interrupting automatic behavior at the moment of decision creates a pause that converts some percentage of harassment into nothing.
None of this fully addresses organized, intentional harassment campaigns. For that, the multi-platform coordination problem remains genuinely unsolved. But for the ambient, normalized toxicity that accounts for most of what marginalized players encounter day-to-day, the design interventions are doing real work.
What You Can Actually Do
I’m going to avoid the instinct to end this with a list of tips, because the framing of “here’s how marginalized players can protect themselves” puts the burden in the wrong place. The burden isn’t on the people being targeted.
What I’ll say instead is directed at the majority of players who aren’t members of the most targeted groups, the ones who are in lobbies when targeting happens and whose response, or non-response, is part of the environmental signal the target receives.
Calling it out when you see it works. Research on bystander intervention consistently shows that one person breaking the silence significantly increases the likelihood that others will too. “That’s not okay” in text chat from one person changes the dynamic of a lobby in a way that silence from five people doesn’t.
The report button, used correctly, specific, timestamped, with a clip where possible, adds signal to the behavioral records that AI moderation systems use. One person seeing and reporting an identity-based harassment incident in their lobby isn’t heroic. It’s just using the tools that exist.
Recommending genuinely welcoming games to friends who’ve had bad experiences matters at a community scale. The games with strong safety reputations, Deep Rock Galactic, FFXIV, Warframe, grew those reputations through player word-of-mouth before the research caught up with the observation. Steering people who’ve been driven out of toxic spaces toward ones with better track records is a real contribution.
And if you make games, work at a studio, or have any connection to the publishing side: 2024 showed clearly that silence in the face of targeted campaigns against developers and players reads as complicity to the people being targeted. That’s worth knowing and weighing.
The Gap Between Tools and Culture
Here’s where I land on all of this.
The AI moderation tools covered in this series are genuinely working at what they’re built to do. Explicit slurs, direct threats, and repeat-offense harassment patterns get caught at rates that would have been impossible three years ago. The data backs this up.
The gap is between what the tools catch and what targeted harassment actually looks like at its worst. Organized campaigns operate across platforms. Dog whistles pass through filters that catch explicit content. The normalization of low-level hostility toward specific groups creates an ambient environment that no individual incident triggers, but accumulates into something that drives people out of gaming entirely.
Closing that gap requires the technology getting more sophisticated, and it is getting there, but it also requires studios being willing to name what’s happening, enforce against it publicly, and signal clearly that the communities they’re building have standards that apply equally regardless of who’s doing the targeting and who’s being targeted.
The studios that have done that, Riot’s hardware bans, Activision’s code of conduct update, the developers who built Deep Rock Galactic into what it is, have built something real. Their communities know it and feel it.
The ones that haven’t yet will keep losing the players they can least afford to lose.
If this article reflects your experience, or doesn’t, I want to hear about it. TAG is actively building our Community Safety Score and community member experiences with safety in specific games are a core part of the methodology. Drop your thoughts in the comments or find me in the TAG Discord.
