The Gendering of A.I.
As Artificial Intelligence technology becomes more mainstream, we frequently come into contact with chatbots and virtual assistants in our lives. We have Siri and Cortana on our phones, Alexa in our homes, and a range of other AIs everywhere we go. While different AIs are produced by different companies and perform different functions, there is one commonality between a large majority of these technological marvels- many happen to display some degree of femininity.
The addition of feminine features to bots is surprisingly prominent in the AI industry, be it Siri’s voice or Sophia’s physical form. Why do companies make their AIs so unquestionably feminine? To answer this question, we must first understand why AIs are so often designed to resemble humans.
Consumers are generally found to be less trusting of AIs than they are of humans. If this lack of trust is not overcome, it will hinder the development of positive human-robot interactions. As a result, the AI may fail to perform the way it is intended to. In order to combat this and improve the consumer experience, companies have resorted to anthropomorphizing their AIs. The results were encouraging. It was found that consumers were more accepting of humanized bots than of non-human ones.
The changes in the AI interface, however, did not end at humanizing the tech. Companies took a step forward and endowed their bots with obviously feminine qualities. Check out the industry and you will find the predominance of ‘female robots’ than male ones. The reason for this biased gendering of AI comes back to the factor of humanness, and more specifically, the ‘perceived’ humanness of both sexes.
Before attempting to understand why there is an overreliance on feminization in the AI industry, one study sought to answer the question- ‘what makes people human.’
While humans have certain characteristics that set them apart from animals and others that distinguish them from machines, there are also characteristics that make some people more human than others. According to Fiske and colleagues, the fundamental dimensions of social judgment are ‘intellect’ and ‘emotion’. We have a tendency to not only categorize humans, but also non-human entities such as AI, on the basis of these two dimensions. We are aware that machines can be quite competent, but where they lack is in the domain of feelings. Turns out, when it comes to AI, consumers want them to have not only competency but also want the essential human characteristic of ‘warmth’.
We perceive men and women as being different when it comes to the domain of emotions. Studies show that women are seen as being ‘more human’ than men because we believe that they are more capable of feeling emotions and expressing warmth. In short, we perceive women as being more emotionally competent than men.
When companies feminize their AI, they are in fact humanizing them one step further than they would if the AI had male or gender-neutral attributes. Of the two social judgment dimensions, warmth is believed to be more characteristic of humans than competence. AI companies, therefore, went on to exploit this gender stereotype of women being more emotional to make their AIs more palatable to their consumers.
Female AI is seen to be more human than its male counterparts. Further, female AIs are assumed to possess more positive human characteristics than male AI, while male AIs are thought to have more negative human attributes than female ones. In addition, the feminizing- and consequent humanizing- of AI makes one assume that the bot will treat them as a unique individual, taking into account their specific conditions (as one expects from humans) rather than being given a cold, standardized treatment.
Not surprisingly, the feminization of AI brings forth an ethical dilemma. Ironically, while feminizing AI humanizes them, it simultaneously dehumanizes real women. The benefits of having AI with feminine characteristics are immense, but it can also be detrimental to the perception of women. According to UNESCO, this trend may further the objectification of women already prevalent in society, painting them as ‘subservient’ and suited for roles that involve serving another. It may reinforce already existing gender stereotypes. With this knowledge in hand, the AI industry must work to maintain a balance in their use of feminizing AI and attempt to find ways to overcome this gender bias in technology so that technological advancement does not come at the price of the well-being of a whole section of society.
References:
Borau, S., Otterbring, T., Laporte, S., & Fosso Wamba, S. (2021). The most human bot: Female gendering increases humanness perceptions of bots and acceptance of ai. Psychology & Marketing. doi:10.1002/mar.21480
