AI in healthcare: How algorithmic bias deepens systemic exclusion in diagnosis and care
Original framing: “Algorithmic gaslighting: The hidden harm of AI in healthcare” — bing news
Marginalized patients, like those in the *#IAmTheAlgorithm* movement, describe AI tools as 'digital gatekeepers' that dismiss their symptoms based on race or gender. A 2022 *BMJ* study found that Black women in the US were 50% more likely to have their pain underestimated by AI triage systems. Trans and non-binary patients report similar erasure, as algorithms trained on binary gender data misclassify their symptoms. These voices are systematically excluded from AI development, reinforcing a cycle where the most vulnerable become the primary victims of 'progress.'
The deployment of AI in healthcare is not a neutral innovation but a continuation of structural inequities, where algorithmic gaslighting becomes the new frontier of medical racism.
The erasure of Indigenous knowledge systems—like *whakapapa*-based diagnostics—mirrors the historical exclusion of marginalized voices from medical research, now automated into bias. Cross-cultural examples, from Japan’s *kansei* AI to Brazil’s community-led telehealth, prove that 'AI' is not a monolith but a tool shaped by power; its failures in the global South expose the myth of its universality. The trickster figure of Eshu reminds us that these systems are not accidents but features of a techno-colonial project, where profit and control trump patient well-being. Solutions must center decolonized data sovereignty, patient-led transparency, and participatory design—otherwise, we risk automating the very biases that have historically denied care to the most vulnerable.