The Code of Bias

As we navigate the technological landscape of April 2026, it has become clear that artificial intelligence is not a neutral force but a mirror reflecting the biases of its creators. Following the "Automating Justice" discussions at the 70th Session of the Commission on the Status of Women (CSW70), a global movement for Feminist AI has emerged to challenge the "algorithmic misogyny" embedded in our digital infrastructure. This movement is grounded in the principles of Data Feminism, which demand that we move beyond simple technical risk management to explicitly examine power asymmetries and make the invisible labor of data production visible. A feminist approach to AI governance requires us to rethink the binaries and hierarchies that current machine learning models reinforce, such as the ways recruitment algorithms continue to penalize "non-traditional" career paths often taken by women or how Large Language Models flatten the context-dependent nature of diverse languages and cultures. In 2026, the call for "Feminist Technology Diplomacy" is gaining momentum, pushing for regulations that move beyond profit-driven engagement metrics—which often amplify extremist and misogynistic content—toward systems that prioritize human dignity and pluralism. We are seeing a shift from "techno-optimism" to a "people-centered" justice approach, where AI tools are audited for gender-responsive outcomes before they are ever deployed into the public sphere. By centering the experiences of those most marginalized by current systems, the Feminist AI movement aims to transform code from a weapon of exclusion into a tool for collective liberation, ensuring that the future of technology is built on a foundation of equity rather than encoded bias.