Anti-Racist Design Principles

See resources for information on how we collaboratively compiled, developed, and deployed these principles. Each project is tagged with its relevant principles.

  1. COUNTERACT: Antiracist technologies recognize and counteract racist technologies or policies by implementing antiracist ones in their stead (i.e. algorithms that confront or change racist policies like gerrymandering)
  2. REDISTRIBUTE: Antiracist technologies analyze and redistribute existing power structures, by redistributing and decentralizing rather than accumulating wealth or power
  3. USERS: Antiracist technologies conceptualize the user with all sorts of identities in mind, including intersectional ones: not just cisgender, white, heterosexual male users. Antiracist technologies are not rooted in gender binaries or heteronormativity. However, antiracist technologies also differentiate between representation and tokenization.
  4. DESIGNERS: Antiracist technologies go beyond representation to make Black and Brown designers integral in the creation process. Antiracist technologies also center women and queer BIPOC as creators. Throughout design work all contributor’s pay cannot be decrimanted because of their ethncity, race, sexuality etc. They offer fair compensation.
  5. COMMUNITY-BUILDING: Antiracist technologies give people a channel to use their antiracist power by offering their time or their resources. They center community over hierarchy. Antiracist technologies support full-stack problem solving. They can help bring people of color together without making it easier to discriminate against, marginalize, or target them.
  6. FLAG: Antiracist technologies recognize, flag, and prevent racist misuse of a platform; and may remove or flags particular racialized technologies from market (i.e. snap filters that make your skin lighter); they inform users about existing policies that maintain or increase existing racial inequalities
  7. NEUTRALITY: Antiracist technologies should not equate whiteness with neutrality. They are not designed for a “neutral” user but are actively designed to include people of color.
  8. DIFFERENCE: Antiracist technologies do not capitalize upon biological differences (i.e. different skin tones) to activate or make a technology work better for one racialized group versus another. As such antiracist technologies act based on the lack of biological differences between all races.
  9. ADDRESSING: Antiracist technologies are those whose purpose is pursued while addressing racial inequality, not neglecting the racism that exists for personal motives
  10. EQUALITY: Antiracist technologies express the idea that racial groups are equals and none needs developing, and support policies that reduces racial inequity. Antiracist technologies do not conflate technological progress for social progress.
  11. BIAS: Antiracist technologies ask users to acknowledge their privileges and power, and where these privileges came from. Antiracist technologies also help people identify their own ethnical racist as ethnic biases can be less apparent and be harder to identify/realize.
  12. CLASS: Antiracist technologies have designs that penetrate disunity between classes and bridge resource/opportunity gaps; they are accessible to people from all socioeconomic groups – affordability and accessibility are not obstacles to anyone. They will not enrich a powerful class at the expense (or even the enrichment) of another.
  13. ACCOUNTABILITY: Antiracist technologies keep leaders, institutions, and people in positions of power accountable; Additionally technological designers and creators understand their accountability and take responsibility for any biased results of the technology they created.
  14. PROFIT: Antiracist technologies are not centered around profit and do not promote profit maximization as a progressive movement. Antiracist technologies change the measure of success from monetary values. They seek to change or avoid “the game” and “go down doing good.” Note that profit may include alternative currencies such as academic publications.
  15. DATA: Antiracist design plays with visibility and invisibility, giving users control over their visibility to others and to opt-out of technical systems. They offer transparancy and agency in how data is used; they do not sell user data and offer the highest protections as the default setting. Being seen/present doesn’t mean you are understood or respected.
  16. DATA BIAS: Antiracist technologies strive to use balanced datasets that are as representative as possible, to weed out bad data that’s biased, or detect biased data that others might be utilizing, and work with all data prioritizing antiracism and equity regardless of how biased or unbiased we perceive it to be.
  17. REDLINING: Antiracist technologies actively undo technological redlining and actively protect social groups when doing any sort of local statewide distribution.
  18. SPEED: Antiracist designers move slowly with purpose. They know that just because you can doesn’t mean you should. They do not neglect ethical or privacy concerns in preference for cheapness or quickness. Evaluation may need longer than just a few weeks to see if a technology is “working”.
  19. SOLUTIONISM: Antiracist technologies surface deep solutions. They recognize nuance and progress and break echo chambers. They do not implement surface-level solutions for deep-rooted issues. Evaluations are ongoing and provide space for users to define how they are using technologies to resolve local problems.
  20. HISTORY: Antiracist technologies investigate and understand the social impact of technological predecessors, actively removing the racist qualities of past iterations
  21. SPACE:  Antiracist technologies must take into consideration racial equity between integrated and protected racialized spaces “which are substantiated by antiracist ideas about racialized spaces.”
  22. EVOLUTION: anti racist technologies are iteratively evaluated to ensure anti racism remains at the core of each development.
  23. TRANSPARENCY: People who develop anti-racist technology must be transparent about their processes (who’s creating the tech, what data is being used, etc.) to ensure that there is an accessible avenue for feedback. This also means clarifying what steps are being taken to make tech more equitable.
  24. JOY: Anti-racist technologies seek and celebrate joy among communities; they aim to “couple critique with creative alternatives that bring to life liberating and joyful ways of living in and organizing our world.” (Benjamin)
  25. CO-LIBERATION: Antiracist technologies do not simply enroll “allies,” but understand that racism affects everyone, such that a push against racism is a push towards the liberation of all.
  26. EXPERTISE: The people we study are experts on their own experience. Designers are not the experts. They should be compensated and respected accordingly.
  27. FAIR TREATMENT. Regardless of who, what, where, whenever we ask someone to provide their expertise or feedback, they must be treated and compensated accordingly.
  28. CONFIGURATION. Anti-racist systems ask us to conduct heterogeneous engineering with new and different communities of people, machines, and objects in the world.
  29. IMAGINARIES. Anti-racist systems center alternative futures produced by non-hegemonic, marginalized groups as possibilities for human-machine configurations.
  30. MUTUAL EVALUATION. The development of anti-racist systems should inspire users and designers to change and grow through a mutual process, and respect expertise on all sides.