About DEI Detector


DEI Detector publishes evidence-based research notes on DEI, woke localisation, and censorship in entertainment media — with a focus on games, movies, TV, anime, manga, and brands/publishers.

What we cover

Standing pages and short reports for individual titles. We look for documented changes and statements, not vibes: trailers, interviews, script comparisons, regional builds, and official marketing.

What counts as woke / DEI / censorship here

  • Agenda pushing — dialogue, characters, or framing rewritten to advance contemporary political or identity messaging rather than the original creative intent.
  • Intentional sexual or feminine censorship in remakes — cuts or redesigns that scrub fanservice, feminine presentation, or similar content when the original (or another region) kept it.
  • Intentional mistranslation / localisation — English (or other) scripts that soften, moralise, or rewrite meaning beyond normal adaptation.
  • Regional censorship notes — when a cleaner or less-edited build exists elsewhere, we call that out.

Woke Score (0–100)

Each researched title can carry a Woke Score from 0 to 100.0 = little or no documented DEI / woke interference;100 = most woke / most DEI. Scores can move as stronger evidence arrives.

How we score

Every score is built from documented evidence. Each page lists the signals it found, what each one adds, and what was checked and ruled out. Rough bands:

  • 0–20: little or no documented DEI or censorship interference.
  • 21–40: minor or isolated signals (a small edit, a debated design choice).
  • 41–60: clear, confirmed signals, but not the core of the work.
  • 61–80: serious signals, such as a credited DEI consultancy or several deliberate changes.
  • 81–100: DEI messaging or censorship is central to the work.

On top of our general criteria above, these rules apply:

  1. DEI / sensitivity consultancies. Confirmed involvement of Sweet Baby Inc. or a similar outright diversity, inclusion or sensitivity consultancy (story or narrative consulting, sensitivity reading, cultural consulting from such a firm) is a serious red flag on its own, whatever it advised. It sets a score floor of about 70, and more if other flags are present. We only count involvement confirmed by credits, official or LinkedIn statements, or reliable reporting. A single unverified curator-list entry isn't enough on its own. Internal "culture trust" or representation review panels that vet characters for identity representation are DEI review by design, so they count the same way, even when staffed by the studio's own employees, once they're confirmed to have shaped that work. Other consultants and advisers (historians, cultural or religious experts, language specialists, community advisers) are judged on what they actually advised. Pure historical, cultural or language accuracy checking isn't penalised. They count only when reliable sources show their advice pushed agenda-aligned or offence-avoidance changes in that specific work, such as sanitising or removing content so a group won't be offended, or adding representation for its own sake. The same applies to internal DEI, HR, sensitivity or content-moderation teams: they count only when sources confirm both what the team is and that it caused such a change. When the evidence of influence is strong but falls short of that, the page carries a "Possible influence (not scored)" note worth 0 points; weak or speculative links get no note.
  2. Track record. Studios and publishers with a documented history of DEI mandates or censorship (for example prior consultancy work, published inclusion mandates, or earlier censored remakes) get less benefit of the doubt on ambiguous changes, plus a penalty of roughly 5 to 15 points. Each page names the documented reason.
  3. Remakes, remasters and re-adaptations. A remake being "modernised" isn't punished on its own. It's a trigger to look closely at what exactly changed. We check the specific content, whether it's woke modernisation or censorship (as opposed to quality-of-life, technical or genuinely artistic changes), who made the change (developer, publisher, platform or ratings board), the stated reason and whether it's credible, and how much it matters. Only woke modernisation or censorship scores, weighted by impact: making a main character less attractive or covering them up, or rewriting their dialogue for political reasons, counts fully, roughly 10 to 20 points; minor or background changes count for little or nothing; unverified claims count for nothing and are listed as unverified. Unexplained changes to important characters lean towards intentional. Where a new work re-creates established characters, we apply the rule at partial weight.
  4. Green flags. Two things lower a score, each by roughly 5 to 10 points: documented attacks or backlash against the work from progressive or feminist critics or outlets, and a developer resisting publisher censorship or moving away from a censoring publisher (for example rejecting censorship publicly, self-publishing or going multiplatform). We cite the evidence and only apply it where it's documented. Fan-forum complaints alone don't count, and neither does criticism that is only about quality. Green flags never erase confirmed red flags from the other rules: pages list both, and we don't count the same evidence twice.
  5. Feminism: which kind. This rule only ever adds points. Traditional strong women, powerful but still feminine, attractive and straight, in the way Japanese anime and games have always written them, are the expected, accepted kind of female strength: they're neutral and score nothing. The modern Western liberal / leftist "girlboss" kind is a red flag, roughly 5 to 15 points depending on how prominent it is: women written as blindly superior or infallible, men written as dumb, useless, disposable or villains for being men, man-hating or "we don't need men" messaging, usually with a deliberately unsexy presentation (think She-Hulk or Ironheart). A female lead on her own is neither.
  6. Male dialogue written out of character. This rule only ever adds points. A line for a male character, especially a strong or gruff one, that sounds like HR, therapy-speak, or a feminine register is a red flag, roughly 5 to 15 points depending on how often it happens and how central the character is. It counts only when that is not the character's established personality and not an intentional design. We quote the line and compare it with how that character has actually spoken. A slightly modern or softer line does not count. A vibe, a rumour, or "people are saying" does not count. A false positive is worse than a low score.
  7. Built-in Pride and identity flag sets. When a developer ships its own dedicated set of Pride or identity flags in the game (emblems, banners, badges or similar), and makes it free and unlocked by default, we treat that as the developer openly supporting the agenda. It scores heavily: a 20-flag set is worth +25, and smaller or larger sets are scaled from there. It counts more when it breaks from a franchise's established tone, such as a gritty war series. We only score a set we or a named source have actually seen in the game, counted flag by flag. A set that is only rumoured scores nothing.
  8. Player-voted LGBTQ+ tag and launch backlash. If the LGBTQ+ tag is verified as a game's #1 player-voted tag on Steam, and there is documented, sourced "woke" backlash in launch week, the two together score about +30. Online reaction counts only when we can name the sources: specific videos, posts or articles. Unsourced "everyone is saying" or "consensus" claims never count.

Regional or government censorship on its own doesn't make the original work woke. When a cleaner build or region exists, we name it.

Standing pages

For a given title we prefer one stable URL. New trailers, interviews, or script dumps get merged into the same page instead of spawning duplicate articles. Check the “last updated” date on each post.

Tags are limited to games, movies, tv, anime,manga, and brand. Browse the blog index or start from thehome page.