The Impact of AI on Women: Progress or Regression in Gender Equality?
As we mark International Women’s Day, it seems like the perfect moment to chat about the complex relationship between women and artificial intelligence (AI), and how it’s reshaping the landscape for women in tech — a realm I’ve been navigating, for better or worse.
AI, the gift that keeps on misgiving.
Let’s imagine AI as that one friend who gifts you a cactus when you’re allergic to plants; it’s the thought that counts, but also, do they even know you? On one side, AI is throwing open doors to new opportunities, promising a tech utopia where everything is sleek, smart, and supposedly equal. Yet, flip the script, and it’s like stepping into a sitcom where gender stereotypes are on repeat, and women are still waiting for their cue to speak.
Take, for instance, those job recruitment algorithms. They seem to have a soft spot for male candidates, leaving women in the digital equivalent of being ghosted after a first date. And our digital assistants, ever so keen to take on the secretary role, reinforcing the idea that women’s “natural” place is in support roles. It’s as if AI’s been snoozing through decades of gender equality progress. These biases aren’t just annoying; they’re harmful, showing women in limited, often sexualised ways and ignoring our broader contributions.
Behind the Silicon curtain.
But the plot thickens, not just in AI’s decision-making but in who’s pulling the strings. The tech world’s gender diversity looks like a pie chart that didn’t quite get the memo on sharing. This leads to all kinds of fun, from health apps that seem to think women’s health is a niche genre to algorithms that believe in recycling stereotypes as if they were going out of style. The logical fix? Ensuring women have seats at the AI table — a no-brainer for creating equitable tech that mirrors our diverse society, yet somehow, it’s as elusive as a stable Wi-Fi connection in a storm.
AI’s impact on women’s employment.
AI’s influence on the workforce presents a paradox for women. On one hand, it’s automating away roles traditionally held by women, such as care work, like it’s on a mission to make us all obsolete. On the other, it’s guarding the gates to STEM fields with a bouncer’s selectiveness, amplifying the digital divide like it’s building a wall, not a bridge.
Women make up 50% of the UK workforce but less than 15% of STEM jobs.
Computing is where the jobs are but women and girls are being left behind.
(Girls who Code)
Take Amazon’s AI recruiting tool, for instance, preferring male candidates as if it was programmed with a “boys only” club mentality, teaching it to overlook female candidates. And then there’s the resume screening tool that nearly skipped over a a highly qualified female candidate for a tech role because it was trained to spot “tech bro” qualities and therefore the algorithm undervalued her profile compared to her male peers, despite comparable qualifications. It’s as if these AI tools are saying, “Diversity? Sounds complicated.”
The algorithmic gender pay gap.
Dive into the realm of finances and employment, where the gender pay gap has sneakily digitised itself, manifesting in algorithms that somehow forget women when showing ads for high-paying jobs. It’s like being left out of a secret club, except the secret is just another rung on the career ladder we’re trying to climb.
Or AI’s influence on the wallet of women. Picture a world where your credit score algorithm is that biased uncle at family gatherings who still thinks women can’t handle money. These AI systems, relying on biased training data leads to algorithms that disadvantage women in areas such as creditworthiness assessment and job suitability evaluation.
Navigating a world designed by men, for men.
Despite the pep talks on gender equality, the tech scene remains predominantly male. Women still face numerous barriers, from stereotypes that we’re not “technical enough”, unequal opportunities and networking events designed either to exclude or lack visible role models.
While initiatives like Women in Machine Learning (WiML) and Girls Who Code are beacons of hope, dismantling the “tech bro” culture, is like trying to clean up glitter: just when you think you’re done, there’s more.
The scarcity of women in tech leadership isn’t just a missed opportunity for diverse brainstorming; it’s like trying to solve a puzzle with half the pieces missing. Elevating women to these roles isn’t just a nice gesture; it’s critical for fostering innovation that reflects and respects the full spectrum of human experience.
The ethical tightrope: Diversity vs. Historical Accuracy.
Then there’s the recent news of Google’s AI model, Gemini, which tried so hard to be inclusive that it ended up rewriting history. This isn’t just about the overt biases; it’s the subtle ones that sneak through, potentially shaping our perceptions of truth. It’s a reminder of the fine line between representing diversity and distorting reality.
Recent headlines have also spotlighted AI-generated images, like those spreading manipulated images of Trump with Black voters, showcasing AI’s power to distort reality and shape perceptions. This misuse of technology not only raises ethical red flags but underscores the urgent need for diverse voices in AI’s creation and regulation to prevent the reinforcement and amplification of biases and stereotypes.
Deepfakes and the commodification of women.
Enter the world of AI imagery, where apps like Midjourney, DALL-E and Stable Diffusion are turning digital art on its head. These tools, while innovative, often reinforce outdated norms showcasing men in roles of intellectual and professional prestige, while women are left to navigate a virtual world cluttered with hyper-sexualised images and an obsession for body types that even Barbie would find unrealistic.
Then there’s Lensa, the avatar-generating app that is turning every woman’s avatar into something out of a questionable fashion magazine, while the men get to be digital CEOs, fully clothed and dignified.
And just when you thought we’d hit peak absurdity, along comes deepfake technology.
Deepfake technology, particularly, presents a chilling frontier. The emergence of apps like ClothOff, as reported by The Guardian, unveils a dark side of AI where women’s images are manipulated without consent, underscoring the urgent need for ethical guidelines and protections in digital content creation.
This digital manipulation extends beyond a violation of privacy; it’s a form of digital violence used to silence, humiliate and blackmail, undermining women’s agency over their own images and contributing to a culture that values women’s appearances over their autonomy.
And as you can all imagine, there is an AI designed specifically for crafting fake nudes. Because, apparently, what the world really needed was more ways to objectify women!
Facial recognition, the misidentification crisis.
Joy Buolamwini’s groundbreaking research on facial recognition technologies revealed their dismal performance in accurately identifying women of colour. They’re nailing it with lighter-skinned men but throw a curveball with darker-skinned women, and it’s like “Error 404: Face Not Found.”
Gender Shades project is a wake-up call. It’s not just about diversity as a nice-to-have; it’s about creating technologies that see and serve everyone, reflecting the true mosaic of human diversity. Unless we’re okay with a future where technology can’t even recognise half the population.
The Digital Divide, when WiFi’s stronger than women’s representation in tech.
And let’s not brush past the digital gender divide, where disparities in access and representation in AI development lead to technologies that echo male perspectives and preferences, sidelining women’s needs and experiences. It’s like voice recognition technologies are saying, “I hear you, but could you drop an octave so that I can understand you more?
This discrepancy isn’t merely inconvenient; it directly impacts women’s ability to interact seamlessly with AI-powered devices in various aspects of daily life, from smartphones to smart home systems.
So how do we flip the script and make AI a place where women don’t just survive — they thrive!
First off, education and awareness. Workshops that don’t just bore you to tears but actually empower you with the AI smarts. Programs like Girls Who Code who are ready to equip the next generation of women with the magic wand of coding to tackle AI’s gender biases head-on.
Next, advocacy and policy influence. It’s like stepping into the arena, but instead of fighting lions, we’re taking on the colossal beast of systemic biases. We’re lobbying for the tech equivalent of the Magna Carta — legislation that ensures every AI system gets a bias audit, kind of like a reality check before it can influence the real world. It’s about holding companies to the fire, ensuring their AI systems are more guardian angels and less rogue robots.
Diverse participation in AI development is the real game changer. Rallying the troops — women and allies, developers, artists, ethicists, and even your grandma — to bring their unique perspectives to the table. Ensuring our tech has a moral compass in its algorithms, that is fair, respects and represents the kaleidoscope of human experience.
Supporting ethically developed AI is like being the eco-friendly, fair-trade consumer of the tech world — good for the soul and the society.
So there you have it, a candid chat on turning the AI world from a boys’ club into a diverse, inclusive, and ethical utopia that not only sees women but values and elevates them. Cheers to the brilliant women in tech facing these challenges with grace, humour, and unstoppable resolve, shaping the future of technology — one algorithm at a time.
Dive Deeper:
AI-Generated Images Targeting Black Voters
Women in Machine Learning (WiML) and Girls Who Code
Gender Shades by Joy Buolamwini
There is no standard’: investigation finds AI algorithms objectify women’s bodies
Women, Not Politicians, Are Targeted Most Often by Deepfake Videos
The viral AI avatar app Lensa undressed me — without my consent
