Misgendering

The act of referring to someone using language that does not reflect their gender identity, such as incorrect pronouns, titles, or gendered terms. In digital accessibility and AI contexts, misgendering occurs when automated systems incorrectly classify a person's gender based on visual or vocal features. This is particularly harmful for transgender, non-binary, and gender non-conforming individuals who may already face frequent misgendering in daily life. Research indicates that automatic misgendering by AI systems can be perceived as even more harmful than human misgendering because it systematically reinforces binary gender assumptions and lacks the capacity for correction through social cues. Assistive technology designers must consider whether gender inference is necessary and how to mitigate harm when errors occur.