Most AI assistants are feminine – and it’s fuelling dangerous stereotypes and abuse
Most AI assistants are feminine – and it’s fuelling dangerous stereotypes and abuse
Publish Date: 2026-01-26 13:54:00
Source Domain: theconversation.com
-
Prevalence of AI Voice Assistants: In 2024, AI voice assistants worldwide surpassed 8 billion, more than one per person on the planet, predominantly with gendered names reinforcing female roles.
-
Gendered Stereotypes in AI Design: The design of AI voice assistants reinforces gender stereotypes: female-named assistants like Apple’s Siri are considered “beautiful women who lead to victory,” whereas IBM’s Watson has a male voice aimed at instructing.
-
Harmful Interactions and Abuse: Research indicates significant instances of verbal and sexual abuse in interactions with feminised AI, with percentages as high as 18% of interactions focusing on sex compared to lower rates for male or non-gendered AI embodiments.
-
Systemic Lack of Regulation and Enforcement: Current regulation fails to address gender bias in AI design; only high-risk systems are reviewed, and enforcement is weak. Many systems ignore gender stereotyping, despite known risks of abuse escalating into real-world behaviour.
-
Technological and Cultural Bias: The lack of female representation in AI development (only 22% globally) contributes to biased designs. The tech culture itself is often sexist, as evidenced by 65% of surveyed senior women in Silicon Valley reporting unwanted sexual advances.
-
Need for Systemic Change: There is a critical need for legislation recognizing gendered harm in AI, enforcing gender-based impact assessments, and demanding accountability. Effective regulation alone isn’t enough; education in tech on the impact of these biases is also crucial.
