New artificial intelligence (AI) technology will now review Disney scripts for gender bias.
The machine learning tool, known as GD-IQ: Spellcheck for Bias, analyses the number of male and female characters in the script to see if it’s representative of the actual population.
Oscar-winning actress Geena Davis formulated the concept through her research-based organisation, the Davis Institute on Gender in Media, with hopes to combat what she describes as “unconscious bias” in Hollywood.
Leveraging patented machine learning technology from the University of Southern California Viterbi School of Engineering, the tool also evaluates the number of LGBTQI characters, people of colour and people with disabilities.
The technology will also assess the number of speaking lines the various groups have, along with the level of sophistication of the vocabulary used and the social status of characters.
Speaking at the Power of Inclusion Summit in New Zealand, Davis announced the partnership with Disney.
“They are our pilot partners and we’re going to collaborate with Disney over the next year using this tool to help their decision-making, identify opportunities to increase diversity and inclusion in the manuscripts that they receive,” she said.
Davis said the program wasn’t designed to “shame and blame” screen creators, but rather to provide data so scripts so don’t perpetuate stereotypes.
“[This is] critically important because the stories we tell in entertainment media can reinforce harmful stereotypes and send a distinct message about who matters most in our culture,” she said.
“We don’t want to change the world; we just want our entertainment to reflect the world as it is – which is half female and incredibly diverse.”
Davis, who had leading roles in Thelma & Louise and The Long Kiss Goodnight, founded her non-profit organisation in 2004 after her growing concerns over lack of equal female representation in movies and television.
The Davis Institute on Gender in Media has since commissioned into how screen representations influence real-world behaviours.