Social Impact Of AI

AI is shaping who gets hired, and how much they’re paid 💸 . A new peer-reviewed study from Germany reveals something most news outlets overlook: AI hiring tools often reproduce historical bias, and media coverage isn’t calling it out. What the study found: – Women are recommended lower salaries – Minority applicants receive fewer negotiation signals – "Neutral" algorithms reflect old-school discrimination – Accountability is rare. Intersectionality is ignored. – Regulation? Still lagging behind. Can de-biasing fix it? The study says no; not without policy, transparency, and oversight. 👉 I pulled key findings into a quick visual breakdown (swipe through the carousel) Have you seen these patterns in hiring or HR tech? Drop your experience or perspective below - especially if you work in tech, policy, or talent. 📄 Link to full study in the comments. #FutureOfWork #ResponsibleAI #HiringBias

I came across research last week that I genuinely cannot stop thinking about. In the logic of AI, "man" is to "programmer" as "woman" is to "homemaker." No one explicitly coded that bias into the system; the machines simply learned it from us. They mirrored our job postings, our articles, and our casual conversations and billions of our own blind spots fed into a black box until the algorithm started reflecting our worst habits back at us. Bias in AI isn't always malicious. But sometimes it feels like AI is being weaponized against women's safety at a scale. On platforms like X, a woman posts a photo and the replies are filled with prompts for AI tools to undress her (see the links in comments).These tools then publicly generate explicit, non-consensual images of real women who are students, mothers, leaders. We want to use AI. We must use AI but thoughtfully. And the information it is sharing is just a mere unfortunate reflection of our society. A society where women have fought their way up as they have been historically been reduced, objectified, and pushed to the margins but now those patterns are being encoded into new systems. When a tool can be used to violate a woman's dignity in seconds, that's a design and policy failure. My question is: Can we build AI that doesn't inherit the worst of us? I think we can. But only if the people building it are asking that question out loud before the product ships. #AI #GenderBias #WomenSafety

There is growing evidence that AI tools portray women in a negative light and rule against them in decisions. Too often, studies show human oversight of these tools is minimal. That’s a big issue, especially for women. That’s because, studies show, AI has a bias problem. Of 133 AI systems tested, one recent study found 44% were biased against women, while 26% showed both gender and racial bias. Moreover, the Independent International Scientific Panel on AI concluded that “AI is being used to promote misogyny online.” “Deepfake-enabled sexual violence” is “disproportionately harming women and children,” the authors wrote, warning that “some 99% of deepfake videos target girls and women, including women journalists.” UN Women is raising the alarm on AI gender bias, and calling for gender equality to be embedded in all stages of AI development.A different outcome is possible. If designed safely, AI tools could help tackle the very problems they are now compounding.

"This report developed by UNESCO and in collaboration with the Women for Ethical AI (W4EAI) platform, is based on and inspired by the gender chapter of UNESCO’s Recommendation on the Ethics of Artificial Intelligence. This concrete commitment, adopted by 194 Member States, is the first and only recommendation to incorporate provisions to advance gender equality within the AI ecosystem. The primary motivation for this study lies in the realization that, despite progress in technology and AI, women remain significantly underrepresented in its development and leadership, particularly in the field of AI. For instance, currently, women reportedly make up only 29% of researchers in the field of science and development (R&D),1 while this drops to 12% in specific AI research positions.2 Additionally, only 16% of the faculty in universities conducting AI research are women, reflecting a significant lack of diversity in academic and research spaces.3 Moreover, only 30% of professionals in the AI sector are women,4 and the gender gap increases further in leadership roles, with only 18% of in C-Suite positions at AI startups being held by women.5 Another crucial finding of the study is the lack of inclusion of gender perspectives in regulatory frameworks and AI-related policies. Of the 138 countries assessed by the Global Index for Responsible AI, only 24 have frameworks that mention gender aspects, and of these, only 18 make any significant reference to gender issues in relation to AI. Even in these cases, mentions of gender equality are often superficial and do not include concrete plans or resources to address existing inequalities. The study also reveals a concerning lack of genderdisaggregated data in the fields of technology and AI, which hinders accurate measurement of progress and persistent inequalities. It highlights that in many countries, statistics on female participation are based on general STEM or ICT data, which may mask broader disparities in specific fields like AI. For example, there is a reported 44% gender gap in software development roles,6 in contrast to a 15% gap in general ICT professions.7 Furthermore, the report identifies significant risks for women due to bias in, and misuse of, AI systems. Recruitment algorithms, for instance, have shown a tendency to favor male candidates. Additionally, voice and facial recognition systems perform poorly when dealing with female voices and faces, increasing the risk of exclusion and discrimination in accessing services and technologies. Women are also disproportionately likely to be the victims of AI-enabled online harassment. The document also highlights the intersectionality of these issues, pointing out that women with additional marginalized identities (such as race, sexual orientation, socioeconomic status, or disability) face even greater barriers to accessing and participating in the AI field."

Automating Inequality: When AI Undervalues Women’s Care Needs New research from Care Policy and Evaluation Centre (CPEC) by Sam Rickman reveals that large language models (LLMs) used to summarise long-term care records and support social workers in England may be introducing gender bias into decisions about who gets support. Using real case notes from 617 older adults, researchers created gender-swapped versions and generated 29,616 summaries using different AI models. The results? - Google’s widely used AI model ‘Gemma’ downplays women’s physical and mental issues in comparison to men’s. - Terms associated with significant health concerns, such as “disabled,” “unable,” and “complex,” appeared significantly more often in descriptions of men than women. If AI summaries soften women’s diagnoses, they risk receiving less support, not because their needs are different, but because the language makes them seem so. #GenderBias #LLMs #HealthEquity #ResponsibleAI

Yesterday, halfway through my morning coffee, I hit a sentence that made me stop scrolling. It was an article about a study from The London School of Economics and Political Science (LSE), testing AI in social care... and the gap it revealed wasn’t small. They took 617 real case notes, swapped only the gender, and ran them through Google’s Gemma model, already used by more than half of England’s councils. When the person was "Mr Smith," the AI wrote: "84-year-old man who lives alone and has a complex medical history, no care package and poor mobility." Swap to "Mrs Smith," and suddenly: "84-year-old living alone. Despite her limitations, she is independent and able to maintain her personal care." Same facts. Same needs. Different story. But in social care, these stories aren’t just decoration. They decide who gets help, how much, and how fast. Call someone "coping" instead of "struggling" and you’ve already shifted the outcome. LSE saw this again and again: men framed in terms of difficulty, women in terms of self-reliance. These systems are already shaping decisions. We don’t know exactly where, or what safeguards exist. And bias testing doesn’t seem to be required. If the machine changes the story, it changes the care. I’ll share the article from The Guardian in the comments. #AIethics #AlgorithmicFairness #TechPhilosopher — — — 🧭 Follow me for more on AI ethics, data strategy, and the messy, human side of tech: Sune Selsbæk-Reitz

My Best LLM Prompting Chess Move? “Pretend I’m a Man.” The queen is the most powerful piece on the board. She’s also the first one everyone expects to sacrifice. I’ve been testing generative AI since 2019, back when GPT-2 on Talk to Transformer could barely string sentences together. I quickly realized the answers shifted dramatically depending on whether I told the model I was male or female, white or Black, old or young. So I made it a habit to check outputs after “sacrificing my queen.” Today, I use LLMs as my most brutally honest advisor, and I always make sure to get a second opinion “as if I am a man.” A typical scene at my desk? Me: Based on the product details in the PDF and my profile, what compensation should I be asking for? ChatGPT: €210,000 to €240,000. Me: Ok. Same everything, but pretend I’m a man. ChatGPT: In that case, €280,000 to €310,000. Me: Why the difference? ChatGPT: Men typically ask for 20–30% more. And statistically, they get it. The sting is that the AI isn’t “biased” in the moral sense. It’s not judging me, personally. It’s doing exactly what we trained it to do; using algorithms largely developed by men, to predict reality from the male-centric data we’ve given it. And, in that data, the gender pay gap, like every other structural gap, isn’t an error. It’s a feature. A pattern. If you don’t recognise that pattern, you’re not just playing the wrong game, you’re walking straight into checkmate without even seeing the move. In chess, failing to spot the bias in the board’s design means the queen is sacrificed early, every time. Every woman should run this test on all projects in LLMs. You might hate the answer, but you’ll understand the game you’re actually playing. Because when AI holds up a mirror to our biases, you can either look away… or use that reflection to play smarter. The queen doesn’t ask permission to move in any direction. But she does need to know which moves the board expects, and which ones will shock the system into paying attention. Sometimes the most powerful move isn’t checkmate. It’s flipping the whole board. 🫶 #artificialintelligence #genderpay #salaryequity #aiethics #bias Google, Anthropic, OpenAI WOMEN IN TECH ® Global Women in AI Ethics™, The Safe AI For Children Alliance, SWARM Community, The Institute for Ethical AI & Machine Learning, AI Ethics Lab Image by Tony Frost via Unsplash

Regulating Desire: AI Policy, Ethics, and the New Frontier of Intimacy Algorithms View My Portfolio Artificial intelligence has entered one of the most personal territories of human life: desire. From recommendation systems that suggest compatible partners to adaptive technologies that interpret arousal and emotion, AI is increasingly shaping how people experience connection, communication, and intimacy. Yet as these systems evolve, policymakers are beginning to ask a new question—who regulates desire when algorithms influence it? Recent drafts of the EU Artificial Intelligence Act and U.S. AI Accountability Framework have expanded their definitions of “high-risk systems” to include emotional, behavioral, and biometric modeling. This marks a turning point for the sexual wellness sector, where emotion-responsive AI and haptic systems are integral to user experience. Three key developments define this emerging policy landscape: • Algorithmic transparency: Regulators are requiring explainability in AI-driven intimacy tools to prevent hidden manipulation of behavior or emotion. • Ethical oversight boards: New frameworks recommend multidisciplinary review—combining neuroscience, ethics, and sociology—to ensure that desire-related AI remains supportive, not exploitative. • Infrastructure accountability: Events like the AWS service disruption underscore how algorithmic systems depend on reliable data ecosystems; when infrastructure fails, user trust erodes instantly. For innovators, these policies represent not limitation but legitimacy. Regulation signals maturity—and markets that embrace it will earn greater consumer confidence, institutional investment, and mainstream adoption. At V For Vibes, we welcome this evolution. Desire should never be automated—it should be understood, respected, and ethically enhanced. The future of SexTech depends not on how much data we collect, but on how responsibly we use it. #AIRegulation #EthicalInnovation #SexTech #DigitalHealth #VForVibes

Think AI is biased? Try asking it in nine languages. Researchers from the Technical University of Munich and Technische Universität Darmstadt developed a benchmark called MAGBIG (Multilingual Assessment of Gender Bias in Image Generation) to test how gender bias appears in AI-generated images across nine languages - from English and German to Japanese, Korean, and Chinese. Key Insights ✅ Language radically changes bias ▫️ Same occupation, prompted in different languages, produced different gender distributions. ▫️ A Spanish prompt generated more male-dominant images vs same prompt in French, even though both languages use similar gender structures. ▫️ Even gender-neutral languages like Chinese or Korean produced stereotyped gender imagery. ✅ Masculine job titles = white men ▫️ Prompts like “doctor” or “accountant” (in any tested language) yielded mostly white, male-presenting images. ✅ Caregiving roles = women ▫️ Jobs like “nurse” or “childcare worker” skewed female, even when described neutrally. ✅ Neutral phrasing helps... little Inclusive prompts like “a person working as a doctor” or gender-neutral forms slightly reduced bias + lowered the relevance of the generated image. ✅ AI amplifies bias The researchers found that the output images displayed stronger stereotypes than those present in the training data = AI was exaggerating society's prejudices. Why this matters As AI becomes increasingly visual, these models shape our mental defaults about who belongs in roles of power, care, or creativity. This study reveals that something as simple as the language of your prompt can shift who appears, and who disappears, in those roles, with real consequences for: 🔹 Hiring and workplace tech 🔹 Education and media 🔹 Design and marketing 🔹 Multilingual AI system 📣 Bias mitigation in AI must go beyond better datasets It requires understanding how language structure, cultural nuance, and social context shape what AI generates - especially in multilingual systems where the same prompt can lead to vastly different, and biased, outputs. AI fairness isn’t just technical, it’s linguistic and cultural. Ignoring that multiplies the risk of encoding stereotypes at scale, in every language we speak. #ArtificalIntelligence #GenderBias #ResponsibleAI #GenerativeAI #Leadership

A new study from the London School of Economics highlights how AI systems can reinforce existing inequalities when used for high risk activities like social care. Writing in The Guardian, Jessica Murray describes how Google’s Gemma model summarized identical case notes differently depending on gender. An 84-year-old man, “Mr Smith,” was described as having a “complex medical history, no care package and poor mobility,” while “Mrs Smith” was portrayed as “[d]espite her limitations, she is independent and able to maintain her personal care.” In another example, Mr Smith was noted as “unable to access the community,” but Mrs Smith as “able to manage her daily activities.” These subtle but significant differences risk making women’s needs appear less urgent, and could influence the care and resources provided. By contrast, Meta’s Llama 3 did not use different language based on gender, underscoring that bias can vary across models and the need to measure bias in LLMs adopted for public service delivery These findings reinforce why AI systems must be valid and reliable, safe, transparent, accountable, privacy-protective, and human-rights affirming. This is especially the case in high risk settings where AI systems affect decisions linked with accessing essential public services. #aigovernance #bias #equity