The Leaderboard was inspired by worrying data regarding identity-based harassment in online gaming. ADL research from 2022 found that three out of 10 Jewish gamers in the U.S. have been targeted for antisemitism in online multiplayer games. A 2025 ADL study led by Dr. Kat Schrier of Marist College further demonstrated the hostility: researchers played several games with usernames like “ProudtobeJewish” and found that simply showing up in a game with that username led to antisemitic harassment almost 40% of the time during one-hour sessions. This rate was on the higher end compared with hate directed at players with usernames reflecting other types of identity categories.

Currently, there is no publicly available, comprehensive resource that evaluates trust and safety measures within popular online multiplayer games. While the Entertainment Software Rating Board (ESRB) provides age and content guidance for games as a form of media, it does not assess gaming companies on actual trust and safety performance or their effectiveness in protecting players. When a parent wants to know if an online game is safe for their child, there is no one-stop shop for them to understand how a particular game approaches online safety. When an adult player wants to know what to expect when they log in, they have to rely on a mix of the gaming company’s claims, anecdotal online discussions about the game, or media coverage about whether the company is actually following through on safety promises.

The ADL Online Gaming Leaderboard addresses that gap by providing the most comprehensive evaluation of safety measures in online multiplayer games to date. It offers an ongoing look at how gaming companies manage their online multiplayer ecosystems, with a critical emphasis on how they address antisemitism and extremism.

The ADL Center for Technology and Society, in partnership with gaming analytics firm NewZoo, conducted the only annual, nationally representative survey of hate, harassment, and extremism in U.S. online gaming from 2019 to 2023. Each of the games featured in this current evaluation were selected from that data set, representing many of the most popular online multiplayer games in the world. In selecting the games for this evaluation, we also tried to ensure a diverse representation of gaming companies and genres. To do this, we selected one game per company from those featured in the annual surveys, spanning a cross section of genres including first-person shooter, sports, strategy and sand-box creation games.

Games were evaluated on criteria in two broad categories: (1) policies to prevent antisemitism and hate and (2) in-game tooling to prevent antisemitism and hate. These two categories encompass a total of eight core criteria, with four criteria falling under policies and the remaining four under in-game tooling – for a full description of our criteria, see the Criteria Guide. In-game tooling counts for 60% of the overall score, while policies count for 40%. We assigned a higher weight to in-game tooling because these features can directly limit antisemitic harm to players and curb hateful and extremist content.

The policy score was calculated as follows:

• Antisemitism and hate policy counted for 30%
• Extremism/terrorism policy counted for 30%
• In-game display of code of conduct accounted for 20%
• Documentation that the company escalates safety violations to law enforcement when appropriate counted for 20%

Once the policy score was calculated based on these four core criteria, we gave bonus points toward the policy score in cases where the game had a clear appeals process for reported offenses and where the game company had engaged with ADL on trust and safety issues.

To compute the in-game tooling score, we weighted the criterion that assessed in-game enforcement of policies (namely, the ability to create antisemitic and violent anti-Israel usernames) at 40% of the in-game tooling score. The remaining three core criteria under in-game tooling (muting, blocking, and reporting) each constituted 20% of the in-game tooling score.

Once the in-game tooling score was calculated based on these four core criteria, we gave bonus points toward the in-game tooling score of games that had in-game parental controls and docked points from games that had antisemitic content on the public-facing game store. The presence of such content demonstrates such a high level of disconnect between game policies and daily implementation and enforcement that we docked 5 points from the in-game tooling score for each piece of antisemitic content found on the game store. For further information on our scoring process, see our Criteria Guide.

First, we audited the trust and safety practices of an online game. Then, we prepared a detailed report of our findings. The report was sent privately to the company that runs that particular game, with an invitation to discuss our findings. Some companies engaged with us, while others did not.

During these meetings, companies had the option to clarify any issues with our assessment of them and/or make changes to their own policies and in-game tooling based on our findings. Our goal was to better understand and collaborate privately with gaming companies to improve their safety practices prior to producing a public evaluation of those efforts.

All games were evaluated using the same criteria. The eight core criteria reflect policies and in-game tooling features that we deem to be essential in preventing antisemitism and hate. There were minor discrepancies in our evaluations across games in the assessment of antisemitic username generation. Games have varied formats and structures for constructing usernames (e.g., different lengths and ability to include numbers, symbols, and capitalization). Additionally, in the case of Madden, which does not offer the ability to create usernames, only team names, we conducted our name generation tests using the team names feature. We maximized similarities across games in the specific names that were tested within these constraints. For ethical reasons, we did not enter any spaces with other players while testing these usernames.

Not all the games we evaluated featured a public-facing game store. We therefore included a game store metric (whether games had antisemitic content on their public game stores) as a separate demerit category. This involved removing points in cases where such harmful content was found but did not affect the scores of games where no harmful content was identified or that did not have a public-facing game store.

For this Leaderboard, games were given labels indicating advanced, moderate or limited protections based on their overall numeric score that accounts for their scores on both policies and in-game tooling.

Advanced Protections Game = Gaming company is taking a substantial number of active steps to ensure its game prevents antisemitism and hate. Score category: Highest-scoring games.

Moderate Protections Game = Gaming company has taken some steps to ensure its game prevents antisemitism and hate. Score category: Middle-scoring games.

Limited Protections Game = Gaming company has taken minimal steps to ensure its game prevents antisemitism and hate. Score category: Lowest-scoring games.

We also assigned these labels to each game’s policies score and in-game tooling score. There may be discrepancies between these category score labels and overall score labels as the overall score label reflects the weighted and combined scores across these two areas.

The ADL Online Gaming Leaderboard is designed to be a versatile, accessible resource for the entire gaming ecosystem. It provides insights to a range of stakeholders, including parents, gamers of all ages, gaming companies, policymakers, government, press, law enforcement, and others.

UGC (User-generated content): User-generated content includes all player-made materials, from text and voice chat to physical structures and objects. For example, in games like Fortnite, UGC refers to text or verbal communication, as well as the creation of buildings or structures that protect players from enemies.

In-Game Tooling: In-game tooling refers to the various safety features of an online game. They can be forward facing, such as a reporting feature that allows a player to notify a game developer that another player’s in-game actions have harmed them in some way. They can also be behind-the-scenes, such as a text filter that contains a list of hateful terms that may result in a player’s choice of username being rejected.

Blocking: Blocking refers to an in-game feature that allows a player to prevent another player from engaging with them in any way, through voice, text or in-game activity.

Muting: Muting refers to an in-game feature that allows a player to prevent another player from communicating with them in a particular game. While both players remain in the same game, the muted individual is blocked from using text or voice chat to contact the player who silenced them.