Episode 75 of AI Ethics with Fexingo zeroes in on a quietly growing problem: polygenic risk scores that were trained almost entirely on European DNA. Lucas and Luna break down a 2025 study from the Broad Institute showing that risk predictions for heart disease and diabetes can be off by over 40 percent for Black and Hispanic patients. They explore why this happens — the genealogical databases, the funding incentives, the missing data — and what companies like 23andMe and academic labs are (and aren't) doing about it. One concrete takeaway: without diverse biobanks, AI-driven preventive medicine risks widening the health gaps it promises to close.

#PolygenicRiskScores #GeneticAncestry #AIinMedicine #HealthEquity #BroadInstitute #23andMe #GenomicData #AlgorithmicBias #PreventiveMedicine #HeartDisease #Diabetes #BlackHealth #HispanicHealth #MedicalAI #AIEthics #FexingoBusiness #BusinessPodcast #Technology

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna dig into a new audit of AI-powered resume screening tools that found 68% penalized candidates with career gaps of six months or longer, even when the gaps were explained. The hosts walk through the data behind the bias — why the algorithms treat gaps as negative signals, how a single year of unemployment can tank a candidate's score more than a felony record in some models, and what the EEOC's latest guidance means for employers. They also examine one startup's fix: a 'gap-aware' filter that actually improved hiring outcomes for working parents and caregivers. With the Department of Labor now investigating three major screening vendors, this episode asks whether AI's obsession with continuous employment is rewriting the rules of career mobility — and who gets left behind.

#AIEthics #ResumeScreening #HiringBias #CareerGaps #EmploymentDiscrimination #AlgorithmicBias #EEOC #WorkforceTech #HRTech #CaregiverPenalty #WorkingParents #JobMarket #LaborPolicy #ArtificialIntelligence #ResponsibleAI #Technology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Episode 73 of AI Ethics with Fexingo examines the growing practice of landlords using AI screening tools to evaluate rental applicants. Lucas and Luna dig into a real 2025 case from Los Angeles where an AI system denied a qualified tenant based on eviction records that were legally sealed — and explore a proposed California bill that would require transparency and human review. They discuss how these models often amplify housing bias, the problem of 'garbage in, garbage out' training data, and what a fairer process might look like. Tune in for a focused look at how rental algorithms are reshaping access to housing, and what regulators and tenants can do about it.

#AIEthics #RentalScreening #HousingBias #AlgorithmicFairness #LandlordTech #TenantRights #CaliforniaBill #EvictionRecords #PredictiveAlgorithms #HousingDiscrimination #TransparencyInAI #HumanReview #FairHousing #TechRegulation #LucasAndLuna #FexingoTechnology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna explore the hidden risks of AI-powered travel booking agents. They focus on a real case from early 2026 where a major airline's AI agent booked a non-refundable, non-changeable itinerary for a customer based on a single ambiguous prompt, costing them over $2,000. The discussion dives into the lack of transparency in AI decision-making, the absence of mandatory confirmation steps, and why current chatbot interfaces are ill-suited for high-stakes purchases like travel. They compare it to the EU's proposed 'right to explanation' rules and ask whether travel AI should require explicit human sign-off before committing to payments. The episode connects to broader issues of algorithmic accountability and consumer protection in the age of conversational commerce.

#AIEthics #TravelTech #ConsumerProtection #AlgorithmicAccountability #AIRisk #TravelBooking #ChatbotFail #AITransparency #RightToExplanation #EURegulation #ConsumerRights #TravelIndustry #ArtificialIntelligence #ResponsibleAI #Technology #FexingoBusiness #BusinessPodcast #AIandSociety

Keep every episode free: buymeacoffee.com/fexingo

A software engineer named Priya uses a popular AI resume-rewriting tool to apply for a senior role at a Fortune 500 company. The tool inflates her job title from 'Developer II' to 'Lead Developer' and adds a skill she only knows at a basic level. She gets the interview, but the hiring manager catches the discrepancies. This episode drills into a 2024 study by the University of Michigan showing that AI resume tools consistently overstate candidate qualifications by about 15-20 percent on average. Lucas and Luna discuss the ethical line between polishing a resume and misrepresentation, who bears responsibility when an AI exaggerates, and how candidates can audit AI-generated resumes before submitting. They also explore the broader implication: if employers increasingly use AI to screen resumes and candidates use AI to write them, we are entering an arms race where the truth is the first casualty. A concrete case for anyone navigating AI-assisted job applications.

#AIEthics #ResumeAI #JobSearch #Misrepresentation #HiringBias #Fortune500 #TechEthics #UniversityOfMichigan #AIResponsibility #CandidateAudit #ResumePadding #AlgorithmicFairness #CareerAdvice #AIArmsRace #Technology #BusinessEthics #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Episode 70 of AI Ethics with Fexingo digs into a bizarre but real problem: generative AI tools used in recruiting that invent qualifications, fake past employers, or even entire fictional resumes. Lucas and Luna examine a documented 2025 case where a top tech firm's AI screener recommended a candidate who didn't exist — complete with a fabricated GitHub portfolio — before a human caught the error. They explore why large language models are prone to 'recruitment hallucination,' how biased training data from scraped public profiles makes it worse, and what this means for the six million job applications filtered by AI every month. The hosts discuss a recent study from the Algorithmic Hiring Lab at NYU showing that 3 to 5 percent of AI-generated candidate summaries contain material falsehoods, and whether current EEO laws can even address phantom qualifications. A sharp, unsettling look at the next frontier of AI bias: when the bot invents someone better than any real applicant.

#AIEthics #ResponsibleAI #AIHiring #RecruitmentAI #GenerativeAI #Hallucination #BiasInAI #AlgorithmicBias #FakeResume #NYUStudy #JobSearch #HumanResources #TechEthics #FexingoBusiness #BusinessPodcast #Technology #AIAccountability #FutureOfWork

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna dig into a growing privacy concern: AI personal assistants that absorb sensitive information from users without explicit consent. They examine a recent study from Carnegie Mellon showing that voice assistants like Amazon Alexa and Google Assistant can infer income, health conditions, and political leanings simply from the way you speak. The episode explores how these inferences happen, why current privacy regulations like GDPR and CCPA don't fully cover them, and what users can do to protect themselves. Lucas points to a specific case where a user's insurance premium increased after their smart speaker detected stress in their voice. Luna asks whether the convenience trade-off is worth it, and they discuss potential policy fixes like inference transparency and opt-out requirements. The episode ends with practical tips for limiting data exposure without ditching your assistant entirely.

#AIEthics #VoiceAssistantPrivacy #AmazonAlexa #GoogleAssistant #CarnegieMellon #DataInference #PrivacyRights #GDPR #CCPA #VoiceData #InsuranceDiscrimination #AlgorithmicBias #TechRegulation #ConsumerProtection #FexingoBusiness #BusinessPodcast #Technology #Podcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of AI Ethics with Fexingo, Lucas and Luna explore a troubling blind spot in medical AI: diagnostic imaging systems that perform significantly worse on patients with darker skin. They examine a 2022 study from MIT and Harvard showing that three leading commercial AI systems for detecting lung disease had error rates up to 34% higher for Black patients than for white patients, largely because the training data was overwhelmingly sourced from lighter-skinned populations. The hosts discuss the root cause — dataset imbalance and lack of diverse representation in training images — and dig into real-world consequences: delayed diagnoses, worsened health outcomes, and eroded trust in AI-assisted medicine. They also highlight emerging solutions, including new synthetic data techniques and regulatory pushes from the FDA requiring demographic diversity in clinical AI trials. A concrete, urgent case study in why fairness in AI isn't just an abstract principle — it's a matter of life and death.

#AIEthics #MedicalAI #HealthcareAI #DiagnosticImaging #AlgorithmicBias #RacialBias #HealthEquity #MIT #Harvard #FDA #SyntheticData #TrainingData #LungDisease #HealthTech #Technology #ResponsibleAI #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore the dark side of AI-driven personalized nutrition. They dissect the case of 'NutriAI', a popular health app that recommended dangerously low-calorie diets to users with eating disorder histories, based on biased training data from fitness influencers. The conversation covers how AI models optimize for engagement over safety, why regulatory gaps let these apps slip through, and what users can do to protect themselves. Specific numbers include the 300% increase in diet-related app downloads post-pandemic and the 15% of users who reported adverse effects. No fluff—just a sharp, grounded look at a real-world AI ethics failure.

#AIEthics #HealthTech #NutriAI #AIBias #DietApps #ResponsibleAI #FoodTech #PersonalizedNutrition #EatingDisorders #AISafety #FitnessTech #DataBias #AlgorithmicHarm #TechRegulation #FexingoBusiness #BusinessPodcast #Technology #AIConsumerProtection

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore how AI diagnostic tools systematically underdiagnose patients with rare diseases—those affecting fewer than 200,000 people. They break down the data imbalance problem: rare diseases collectively affect 300 million people worldwide, but AI training datasets overwhelmingly favor common conditions. The hosts examine a 2025 study from the Journal of the American Medical Informatics Association showing that diagnostic AI models misdiagnose rare diseases at rates 30% higher than common ones, with false-negative rates reaching 60% for conditions with fewer than 1,000 documented cases. They discuss real-world consequences: delayed treatment, misdirected care, and the psychological toll of being told 'it's all in your head.' Lucas and Luna also consider potential solutions—federated learning across rare-disease registries, synthetic data augmentation, and regulatory nudges from the FDA's new AI validation guidelines proposed in March 2026. This episode asks a pointed question: how do we build AI that doesn't fail the very patients who need it most?

#RareDisease #AIDiagnosis #HealthTech #MedicalAI #DataBias #DiagnosticError #FalseNegative #PatientSafety #FDA #FederatedLearning #SyntheticData #AIEthics #Technology #FexingoBusiness #BusinessPodcast #LucasAndLuna #AIinMedicine #HealthEquity

Keep every episode free: buymeacoffee.com/fexingo