AI for Diversity and Inclusion Assessments

𝗔𝗜 𝗶𝘀 𝗰𝗵𝗮𝗻𝗴𝗶𝗻𝗴 𝗵𝗼𝘄 𝘄𝗲 𝗵𝗶𝗿𝗲, 𝗯𝘂𝘁 𝗶𝘀 𝗶𝘁 𝗵𝗲𝗹𝗽𝗶𝗻𝗴 𝘂𝘀 𝗯𝘂𝗶𝗹𝗱 𝗶𝗻𝗰𝗹𝘂𝘀𝗶𝘃𝗲 𝘁𝗲𝗮𝗺𝘀? Many companies I partner with, including Fortune 500s, have started using AI for hiring from resume screening to video interviews. And I'm a big advocate for these tools because they help us hire faster and more fairly. But here's what many may not realise. It's not about just using AI. It's about using it the right way. This is really important because that ensures that candidates are all truly assessed for their skills. So if you are wanting to build an inclusive hiring process with AI, here are 5 ways to get started: 1️⃣ 𝗗𝗲𝗳𝗶𝗻𝗲 𝘆𝗼𝘂𝗿 𝗴𝗼𝗮𝗹𝘀. Get clear about what fair and inclusive hiring means for your team before adding AI. This way you'll have clear measures of success too. What gets measured, gets tracked. 2️⃣ 𝗨𝘀𝗲 𝗱𝗶𝘃𝗲𝗿𝘀𝗲 𝗱𝗮𝘁𝗮. We all know that AI is only as good as what we feed it. To give yourself the best chances, make sure your data input reflects real diversity, across race, gender, age, ability, and more. 3️⃣ 𝗣𝗮𝗿𝘁𝗻𝗲𝗿 𝘄𝗶𝘁𝗵 𝘆𝗼𝘂𝗿 𝘁𝗲𝗰𝗵 𝘃𝗲𝗻𝗱𝗼𝗿𝘀 Ask how they test for bias and what proof they have their tools are fair. Inclusion is a shared responsibility. 4️⃣ 𝗖𝗵𝗲𝗰𝗸 𝘆𝗼𝘂𝗿 𝗿𝗲𝘀𝘂𝗹𝘁𝘀 𝗼𝗳𝘁𝗲𝗻. I always say, everything is data. Look for patterns in who gets filtered out. One client found their AI was missing career changers—something we only caught by reviewing the data. 5️⃣ 𝗞𝗲𝗲𝗽 𝗽𝗲𝗼𝗽𝗹𝗲 𝗶𝗻𝘃𝗼𝗹𝘃𝗲𝗱. Yes, AI helps, but we still need humans making the big calls. Train your team to spot and correct bias, whether it comes from tech or people. What are some inclusive hiring practices you've seen? I'd love to hear your stories! #inclusivehiring #airecruitment #lfbalumni #diversityandinclusion  

Why Transgender and Gender expansive Representation Matters in AI/ML ▶️ Bias in the Dataset = Bias in the Outcome: If transgender data is absent or erased, AI systems will assume “cisgender” as the default. That leads to flawed outputs: misgendering in chatbots, errors in health risk predictions, exclusion in credit scoring, or discriminatory hiring algorithms. ▶️ Healthcare Impact: Without transgender-specific health data, predictive models may ignore conditions tied to hormone therapy, surgical outcomes, or mental health disparities. This results in poorer care and systemic inequities in clinical decision support systems. ▶️ Social & Economic Impact: Credit and risk-scoring systems often misinterpret transgender people’s data (e.g., name changes, gaps in work history, or mismatched IDs). These errors can block fair access to loans, jobs, or housing. Beyond data points, trans experiences bring unique perspectives about resilience, adaptation, and identity. These should inform how models are designed to respect human dignity and diversity.

Goldman Sachs just dropped its mandate requiring diverse board members for IPOs. As major companies like Amazon, Meta, and Goldman Sachs reshape their DEI approaches, new AI ALPI research reveals a striking reality: Companies that maintain strategic DEI initiatives are quietly building unprecedented competitive advantages. The Current Landscape → Major tech giants reconsidering DEI programs → Increased regulatory scrutiny of traditional approaches → Shift from quotas to performance-driven inclusion But Here's What the Data Actually Shows: 1. Performance Metrics That Can't Be Ignored → 39% higher financial performance with diverse executive teams → 87% better decision-making in diverse environments → 70% higher likelihood of capturing new markets → 30% performance boost in high-diversity settings 2. The AI-Powered Evolution Instead of retreating, leading organizations are revolutionizing DEI through AI: → Eliminating bias in hiring through objective data analysis → Creating personalized development paths at scale → Measuring inclusion impact in real-time → Predicting and preventing equity gaps before they emerge 3. The New Strategic Imperative Smart companies aren't choosing sides in the political debate – they're: → Moving from compliance to competitive advantage → Leveraging AI for merit-based, bias-free decisions → Building inclusive cultures that drive innovation → Measuring DEI impact on business performance The Market Reality: While headlines focus on companies scaling back, industry leaders are quietly transforming DEI into a data-driven performance engine. The gap between innovators and laggards is widening. Implementation Framework for 2025 → Deploy AI-powered analytics for objective decision-making → Focus on measurable business outcomes → Build cross-functional transformation teams → Create sustainable, technology-enabled processes The future of DEI isn't about politics – it's about performance. Organizations using AI to drive inclusive excellence are seeing unprecedented returns on both talent and business metrics. 🔥 Want more breakdowns like this? Follow along for insights on: → Getting started with AI in HR teams → Scaling AI adoption across HR functions → Building AI competency in HR departments → Taking HR AI platforms to enterprise market → Developing HR AI products that solve real problems #FutureOfWork #DEI #AIinHR #LeadershipStrategy #WorkplaceCulture #HRTech #Innovation

Most frameworks count how many people with disabilitied are in the room. Ours measures what’s keeping them out. There’s a fundamental difference and it changes everything about what we do next. Organisations celebrate diversity numbers while the environment around people with disabilities stay exactly the same. Inaccessible. Exclusionary. Unchanged. A Monitoring & Evaluation Framework that doesn’t ask “are people with disabilities included?” It asks: → What barriers exist in this environment? → How deep does the exclusion run? → What has persisted and why? This is the social model applied to measurement. We stop locating the problem in the person. We locate it in the system. Because you can’t fix what you’re not accurately tracking. This framework is the foundation of how AccessInclusion conducts accessibility assessments and it’s reshaping how our clients understand what inclusion actually requires of them. Remediation without rigorous measurement is just guesswork with good intentions. We deserve better than that. I’m building out the methodology white paper. If you work in disability inclusion, policy, or organisational M&E and you’re tired of metrics that celebrate presence without interrogating barriers let’s talk #DisabilityInclusion #MonitoringAndEvaluation #InclusionFramework #AccessInclusionInstitute

Feel like you have the perfect dataset to train an inclusive AI model? Think again. Have you truly considered someone's full, embodied experience beyond the digital footprint that exists about them? ❓ "𝗗𝗶𝗴𝗶𝘁𝗮𝗹 𝗱𝗮𝘁𝗮 𝘁𝗿𝗮𝗶𝗹𝘀 𝘁𝗲𝗹𝗹 𝗮 𝘀𝘁𝗼𝗿𝘆, 𝗯𝘂𝘁 𝗶𝘀 𝗶𝘁 𝘁𝗵𝗲 𝗿𝗶𝗴𝗵𝘁 𝘀𝘁𝗼𝗿𝘆?" ❓ Often, we build training datasets based on available data. Unfortunately, this reifies existing digital inequalities as some people have more digital traces than others. 𝗧𝗼 𝗯𝘂𝗶𝗹𝗱 𝗺𝗼𝗿𝗲 𝗶𝗻𝗰𝗹𝘂𝘀𝗶𝘃𝗲 𝗔𝗜 𝗺𝘆 𝗿𝗲𝗰𝗼𝗺𝗺𝗲𝗻𝗱𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗿𝗲: 1. Start by assessing your dataset for who is, and who is not, included. 2. Think about the lives of the people included, build 5 “day in the life” personas. 3. Assess each “day in the life” for what data points that are digitally captured. 𝘼𝙣𝙙 𝙬𝙝𝙖𝙩 𝙩𝙝𝙚𝙮 𝙢𝙞𝙨𝙨.  4. Groundstruth your findings with your end users, asking them to fill in any blanks and validate your assessment. 5. Refine your data sources to more holistically capture the realities of the people you aim to serve. Or, gulp, walk away from AI for now. (yes, that's an option). If you're interested in inclusive AI, I recommend reading the amazing work of Alexandra R. , Alex Kessler, and Jacobo Menajovsky through Center for Financial Inclusion (CFI).                                                        Read the full report: https://lnkd.in/eNwFW4Ha

New responsible AI paper with Christina N. Harrington, PhD and Shaun Kane! We synthesized categories of disability, health, and accessibility representation and intersectional harms, as evaluated by people with diverse disabilities and health conditions who created AI images with us during an interactive interview study. We'll present this at the upcoming Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO) conference! Takeaways: * Disability representation is about so much more than disabilities; not everyone who experiences health conditions or ableism identifies as disabled. Our expansion to disability, health, and accessibility (DHA) was intentional and is in deference to Sami Schalk's book, Black Disability Politics and the greater Disability Justice Movement. * DHA representation in AI images not only refers to how people look but also concerns access technologies, objects, actions, and motions that may signify symptoms, and when combined with generic terms like activities, participants expected that they be shown done in an accessible manner (e.g., an image of disabled people doing yoga should show adaptive yoga techniques). * We point out intersectional harms including: environments were often depicted as upper class, colorism (in addition to predominance of default AI images of white disabled people), and body size homogenization. * Often, successive prompts did not result in "better" images, but could even be worse than original images--something our participants identified particularly when they prompted for multiple representation characteristics (e.g., people with different races, ages, and disabilities). What do we do now? * Proactively include DHA in evals: Models change but people with disabilities and health conditions deserve access to realistic representations which dignify and celebrate diverse life experiences and identities. If you can make better AI image representations than this paper, great. We still need to evaluate representation inclusive of DHA; this need will not go away. * Community members need not just evaluate outputs, but should develop test prompts in their own words. * Our categories could scaffold term taxonomies and test prompt sets. * Intersectional harms is *another* reason for those of us focusing on different aspects of representation to work together. * Iteratively prompting AI models was an interesting method to spur evaluations, warranting future research. * Accessible research is not just for accessibility research: I and others have said this but it is always worth mentioning. The substance of evaluations doesn't matter if the tools, environments, and processes for eval completion aren't accessible. https://lnkd.in/ehsQ_cAd

AI is only as inclusive as the voices driving its development. The way we build and implement AI today will determine how it serves tomorrow. The choice is ours. It has the potential to reshape industries, but if left unchecked, it risks deepening societal divides and the inclusion gap. While we see progress. Western-centric AI development has perpetuated biases by relying on incomplete data and overlooking underserved regions. To shift this narrative, we need to move beyond the buzzwords and focus on tangible actions. Here’s how: → Diversify the data: We must actively collect and incorporate data from underrepresented regions, ensuring AI systems reflect diverse needs and experiences. → Empower diverse talent: AI development must include voices from all communities. We need initiatives that nurture talent in underserved populations to bring fresh perspectives into tech. → Engage globally: Policymakers, tech companies, and healthcare providers must collaborate, ensuring AI solutions are designed for global accessibility. → Hold ourselves accountable: Regular audits for bias in AI systems should become the norm. → Rethink governance: We need inclusive AI governance that prioritizes representation, particularly when it comes to health and social welfare. → Learn from local experts: Before implementing AI in new regions, tech developers must work alongside local experts to understand cultural nuances and real-world needs. Moreover, applying the 4D Framework: Develop, De-identify, Decipher, De-bias. We can create AI systems that are not just smarter, but also fairer, more inclusive, and global. It’s time to change the conversation. But this isn't just about building better tech. It's about expanding access, education, and funding to communities that have been left behind. It’s about ensuring that every person, no matter where they live, has a seat at the table. AI’s future doesn’t belong to one group. It belongs to all of us. The real question is: Will we design it for everyone?

🚫 Want your AI to exclude customers or users? Focus only on testing for, and preventing, bias in your data, model, and output. ⚠️ Obviously you want your AI to be as inclusive as possible. Inclusive AI ensures greater market share, prevents harm, and bolsters your reputation as a responsible and trustworthy industry player, so you need to also consider potential bias in your workforce (and yourself). ❓who built, labelled, and sourced the data, and how could their biases be embedded in the data and flow through the model❓ ❓who makes decisions. How, and by whom, are they informed❓ ❓have you sufficiently multidisciplinary teams with different lived experiences to bring new types of thinking and drive better outcomes❓ ❓who is able to provide feedback and how, inclusive of both employees and your consumers. Who listens to, and decides what feedback is acted upon❓ ❓what training, accountability, and governance have you implemented to ensure employees are equipped to identify and address potential harmful biases❓

Sony Establishes a New Global Standard for Ethical, Bias-Aware AI Evaluation Introduction Sony AI has launched a breakthrough dataset designed to expose and measure bias in computer vision systems. The Fair Human-Centric Image Benchmark (FHIBE) represents a new ethical baseline for AI development, emphasizing consent-driven data collection and global diversity—two elements long missing in mainstream AI training pipelines. Key Developments • FHIBE is the first publicly available, fully consent-based, globally diverse human image dataset built explicitly to test fairness in AI models. • The dataset includes images of nearly 2,000 paid participants from more than 80 countries, each with the option to remove their images at any time. • Images carry detailed annotations covering demographics, physical characteristics, environmental context, and camera metadata—allowing precise diagnosis of bias sources. • Sony found that no existing dataset from any major company met FHIBE’s fairness criteria. • Testing revealed that current AI models still exhibit deeply entrenched biases, including lower accuracy for individuals using she/her pronouns driven in part by overlooked hairstyle variation. • When prompted with neutral questions about occupation, models frequently produced harmful stereotypes, particularly against specific pronoun groups and people of African or Asian ancestry. • Some models disproportionately generated toxic or criminalizing descriptions for darker-skinned individuals and those identifying with he/him pronouns. • FHIBE validated well-known biases but also uncovered new contributing factors, showcasing its power as a diagnostic framework. Broader Context and Implications • FHIBE demonstrates that ethical data collection—transparent, consent-based, geographically inclusive—is not only feasible but necessary for trustworthy AI. • The benchmark gives researchers, companies, and regulators a standardized tool to evaluate fairness across vision tasks, from identification and classification to generative modeling. • Its release signals growing industry pressure to replace web-scraped datasets with ethically sourced alternatives. • Sony plans ongoing updates, forming a dynamic, living resource as AI systems evolve and fairness standards advance. Why This Matters AI bias is no longer an abstract concern—it shapes real-world outcomes in hiring, healthcare, law enforcement, and digital platforms. FHIBE provides a tangible path toward accountable AI development by exposing harmful patterns and creating a shared baseline for improvement. Sony’s work underscores a pivotal truth: fairness is not a byproduct of scale; it is the result of intentional, ethical engineering. FHIBE offers the industry a credible route to build systems that treat people with accuracy, dignity, and equity. If this topic resonates, I invite you to connect and continue the conversation. Keith King https://lnkd.in/gHPvUttw

Why have a Diversity Dashboard of your company? Leveraging a technology platform like UNGENDER Advisory that provides a diversity dashboard can enable companies to make better decisions regarding team composition, hiring, promotions, and company culture. Here are some more reasons why and how companies can make the most of such a platform: Objective assessments: Remove biases and subjectivity by relying on data-driven metrics. Having quantifiable data on the diversity of teams allows for objective evaluations of current diversity initiatives and areas of improvement. Talent acquisition and retention: Understanding the diversity breakdown can help in targeting hiring campaigns more effectively. If certain groups are underrepresented, recruitment efforts can be tailored accordingly. Additionally, seeing patterns of attrition among certain demographics might indicate issues to address in the work environment. Setting clear targets: Establish diversity and inclusion benchmarks based on the data presented. Targets can be based on industry averages, competitors, or a company’s own aspirations. With clear metrics to track, progress can be measured over time. Identifying training needs: If there's a lack of diversity at the leadership levels, companies might consider offering training programs, mentorship opportunities, or pathways to leadership for underrepresented groups. Feedback mechanism: Use the platform as a way to get feedback from employees. If there are specific concerns or feedback related to diversity, the platform might offer ways to anonymously submit these concerns. Tailored initiatives: By understanding which groups are underrepresented, companies can launch specific initiatives, like women in tech programs, LGBTQ+ inclusivity training, etc., tailored to their unique needs. Promoting inclusive culture: Use the insights from the dashboard to reinforce the importance of diversity and inclusion in the company's culture. This can lead to more empathetic and collaborative work environments. Stakeholder communication: Share the findings with stakeholders, including shareholders, customers, and employees. Transparency regarding diversity can enhance a company's reputation and show a commitment to better decision-making. Decision making: Having a diversity dashboard isn’t just about tracking; it’s also about incorporating those insights into actual decision-making. Whether it’s related to product development, marketing campaigns, or strategic planning, a diverse team offers varied perspectives, leading to more comprehensive and innovative solutions. Regular reviews: Continuously monitor the dashboard and revisit decisions, targets, and initiatives regularly. This ensures that the company remains proactive in its efforts and adjusts strategies based on real-time feedback and changes. Want to know more? Write to us at contact@ungender.in or visit www.ungender.in to schedule a demo.