Systemic gender bias and undervalued care work make female-dominated occupations the most vulnerable to AI automation
Original framing: “These 3 charts show female-dominated jobs are actually the most exposed to AI” — The Conversation - Global
Empirical studies using occupation‑level automation risk indices consistently find that jobs with high interpersonal interaction and emotional labour—often female‑dominated—are flagged as high‑risk due to advances in natural language processing and affective computing. However, methodological biases arise from training data that underrepresent care work, inflating risk estimates. Robust, interdisciplinary research is required to refine these models.
The vulnerability of female‑dominated occupations to AI is not an accidental side‑effect but the outcome of a century‑long pattern where market‑driven automation targets undervalued care work, reinforced by gendered data biases and absent policy safeguards.
Historical precedents, Indigenous relational ethics, and cross‑cultural welfare models demonstrate that alternative trajectories are possible when power is redistributed through gender‑responsive governance, public investment in care‑centric skills, and cooperative ownership of AI. By foregrounding marginalised voices and employing trickster insights, we can reframe AI from a deterministic threat into a lever for restructuring labour relations toward equity and dignity. Immediate action by states, unions, and community organisations is required to embed these systemic solutions before the next wave of automation entrenches existing inequities.