Gabriel Geiger @gabrielgeiger.bsky.social · 30/06/2026Was such a pleasure to speak with @alixdunn.com about automated governance alongside this stellar lineup 032
Reposted by Gabriel GeigerThe Maybe @themaybe.org · 30/06/2026Governments across the Global Majority are trying to tackle tough problems with tech. But instead of creating the oft-promised efficiencies, these tools are delivering deeper inequities–while making them harder to see and challenge. 132
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026It was a pleasure working with the @wired.com @the-independent.com teams, including @regret.bsky.social @mattburgess1.bsky.social, @couts.bsky.social @hollybancroft.bsky.social 010
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026We found that it classifies more than a 1/3 minors as adults in the best case and more than 2/3 in others. Furthermore, we found that the system was more likely to classify minors from West Africa as adults compared to Europeans. Read our methodology www.lighthousereports.com/methodology/...lighthousereports.comHow We Analyzed AI Used to Guess Asylum Seekers’ AgesUsing a leaked report and NIST data to audit facial age estimation AI for efficacy and bias. 101
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026We later found the vendor of the FAE tool, a company called Cognitec. NIST conducts assessments of FAE vendors including Cognitec and publishes the results. We extracted the underlying data and conducted an audit. 100
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026The average error for female Sub-Saharan Africans was more than 4.5 years, meaning a 13-year-old could be classified as an adult. @the-independent.com www.independent.co.uk/news/uk/home...independent.co.ukHome Office knew AI age checks for migrants were flawed — but rolled it out anywayExclusive: Government advisors describe the tool as ‘hideously flawed’ amid concern facial age estimation has ‘baked-in racial bias’ — and can wrongly identify a 14-year-old as an adult 100
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026But she obtained a damming internal Home Office report that showed how the “best performing system” was biased against Sub-Saharan Africans, who were the largest group arriving on small boats in 2025 100
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026Earlier this year, @maybulman.bsky.social caught wind of the Home Office’s plans to deploy FAE to assesses asylum seekers ages, which they claimed would “root out illegal migrants gaming the system by posing as children.” 100
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026It's hard to imagine a decision with greater stakes. Yet facial age estimation technology (FAE) is mainly used to stop children from buying booze and cigarettes. It’s now being repurposed to make life-changing decisions about children at their most vulnerable @wired.com www.wired.com/story/facial...wired.comThe UK Will Scan Asylum-Seekers’ Faces for Age Checks—Despite Knowing the Tech Is FlawedInternal Home Office tests of age-verification technology show the risks of life-altering errors. It’s moving forward anyway. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026And that's a huge problem, because that determination has profound consequences, affecting everything from the rights they are afforded to the accommodation they receive while their claim is processed. 100
Gabriel Geiger @gabrielgeiger.bsky.social · 18/06/2026Soon, when a 15-year-old asylum seeker arrives in the UK, there’s a good chance that a border officer will feed a picture of their face into an AI age estimation tool that will guess whether they are an adult or not. New @lighthousereports.com investigation shows that tool is deeply flawed 1911
Reposted by Gabriel GeigerAntonella Napolitano @svaroschi.bsky.social · 18/06/2026Incredibly important story by @maybulman.bsky.social and team at @lighthousereports.com with WIRED and The Independent. @gabrielgeiger.bsky.social @mattburgess1.bsky.social @hollybancroft.bsky.social Maddy Varner Sadly, the shape of things to come in the EU www.lighthousereports.com/investigatio...lighthousereports.comAsylum by AlgorithmThe UK has turned to AI to determine the age of asylum seekers, and other countries are likely to follow suit. Our investigation finds the technology is deeply unreliable. 122
Gabriel Geiger @gabrielgeiger.bsky.social · 07/05/2026It was such a pleasure to speak with @themaybe.org about our new investigation 120
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026This has been an enormous team effort. Find all the resulting publications, including a technical methodology here. www.lighthousereports.com/investigatio...lighthousereports.comHiding Behind AIHow Kenya Used a Predictive Algorithm to Transfer Health Costs onto the Poorest 030
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026Another person told us about how her premium meant she had to choose between feeding her children or paying for healthcare. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026The consequences of this on the ground are almost indescribable. A frontline health worker told us people dying because they couldn’t access essential care after being charged premiums that were too high. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026What’s more, a confidential document we obtained showed that the government knew that its model was “inequitable, especially for poor households” but deployed the system regardless. www.theguardian.com/global-devel...theguardian.comFlaws in Kenya’s AI-driven health reforms driving up costs for the poorestExclusive: amid unrest, President William Ruto promised to give all Kenyans access to healthcare. But the algorithm favours the rich, an investigation has found 141
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026It’s remarkably crude, using variables like whether your walls are made of dung or plywood, you own a radio, and your level of education to predict your income. Simply having access to electricity, or completing high school can lead to an actually poor household predicted as rich. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026What we found was striking: the system overcharges the poorest Kenyans while undercharging the wealthiest. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026With that, we reconstructed the model and audited it, examining how it calculates premiums for actual Kenyan households. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026To truly understand the system, we had to see inside it. Using access-to-information laws and confidential sources, we secured the training data and underlying formula, as well as a detailed record of how it was built. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026Yet the government continued to claim that the system was objective and fair. The system was presented as a panacea for achieving universal healthcare, including for the poorest Kenyans. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026Early on, we saw signs that this system wasn’t working. Thousands of Kenyans went online to complain about AI-calculated healthcare premiums they could not understand or afford. www.youtube.com/watch?v=2ylE...youtube.comERROR BY DESIGN: How SHA's AI system is failing KenyansYouTube video by Africa Uncensored 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026The centrepiece is an algorithm that predicts how much money the country’s informal workers, the 80 percent of the country who do not have formal income records, earn. They are then charged a proportion of their predicted income. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026The Kenyan government is in the midst of a sweeping experiment in using artificial intelligence to determine how much people should pay to access public healthcare. 110
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026A team of us at @africauncensored.bsky.social @lighthousereports.com @theguardian.com have spent months trying to unravel how this system works. sha.africauncensored.onlinesha.africauncensored.onlineHiding Behind AIHow SHA Was Used to Load Health System Costs Onto Poorest. 141
Gabriel Geiger @gabrielgeiger.bsky.social · 04/05/2026Kenya is using AI that assesses your toilet to set your healthcare bill. I’ve reported on this topic for a long time, across many different countries. None have struck me as this consequential, with stakes so high. 142
Reposted by Gabriel GeigerGlobal Investigative Journalism Network @gijn.org · 24/03/2026In this masterclass video, part of @gijn.org's new #TechFocus, Lighthouse Reports investigative journalist Gabriel Geiger shares practical tips for reporting on how governments use algorithms and AI systems for policing, welfare, & healthcare. Watch the full video ⬇️ twp.ai/E6IByA 086
Gabriel Geiger @gabrielgeiger.bsky.social · 01/01/2026Can confirm ditching flourish. I know I need to atone for my sins 120
Reposted by Gabriel GeigerSimon Munzert @simonsaysnothin.bsky.social · 20/12/2025🎅2️⃣0️⃣ Inside the Suspicion Machine, by @evaconstantaras.bsky.social, @gabrielgeiger.bsky.social, @jusbraun.bsky.social, @dmehro.bsky.social, and Htet Aung from @lighthousereports.com in @wired.com www.wired.com/story/welfar...wired.comInside the Suspicion MachineObscure government algorithms are making life-changing decisions about millions of people around the world. Here, for the first time, we reveal how one of these systems works. 143
Reposted by Gabriel GeigerThe Syllabus @the-syllabus.bsky.social · 26/10/2025Unveiling the murky underbelly of the surveillance industry through the lens of First Wap, this investigation critiques how the company navigates regulatory frameworks to sell powerful surveillance tools. By @gabrielgeiger.bsky.social, @lighthousereports.com in @motherjones.combuff.lyThe surveillance empire that tracked world leaders, a Vatican enemy, and maybe youInside the hidden world of First Wap, whose untraceable tech has targeted politicians, journalists, celebrities, and activists around the globe. 053
Reposted by Gabriel GeigerDr Abeba Birhane @abeba.blacksky.app · 21/10/2025"The archive contained more than a million tracking operations: efforts to grab real-time locations of thousands of people worldwide. What emerged is one of the most complete pictures to date of the modern surveillance industry." 0269
Gabriel Geiger @gabrielgeiger.bsky.social · 19/10/2025You all did an epic job putting it together 020
Reposted by Gabriel GeigerLu Olkowski @luolkowski.bsky.social · 19/10/2025What I’ve been working on… Incredible reporting from @gabrielgeiger.bsky.social and @cr0ft0n.bsky.social at @lighthousereports.com Then, Michael Montgomery @revealnews.org made it into a radio story that I was lucky enough to edit. What a partnership. 14 newsrooms. 😮 35822
Gabriel Geiger @gabrielgeiger.bsky.social · 19/10/2025Incredible to work with the @revealnews.org team on this one, including Michael Montgomery , @luolkowski.bsky.social and @artiscuriskis.bsky.social 171
Gabriel Geiger @gabrielgeiger.bsky.social · 19/10/2025You’ll also hear how First Wap’s executives tracked their own decides as they crisscrossed the globe on an authoritarian sales tour. @rondeibert.bsky.social of @citizenlab.ca describes this as “despotism as a service.” 171
Gabriel Geiger @gabrielgeiger.bsky.social · 19/10/2025Hear from a journalist surveilled while investigating the Vatican, a woman tracked by her stalker, and the daughter of a Rwandan opposition leader targeted in the lead up to his murder. 181
Gabriel Geiger @gabrielgeiger.bsky.social · 19/10/2025We made more than 100 cold calls. Each one began the same way: “You’ve been targeted by government-grade surveillance tech.” In our @lighthousereports.com collab with @revealnews.org, you can hear the voices on the other end. revealnews.org/podcast/cell...revealnews.orgA young reporter discovers a mysterious trove of data that exposes a global surveillance empire.In a major investigation, a young reporter uncovers a powerful technology used to spy on thousands of people across the world. 25827
Reposted by Gabriel GeigerLaurin Lorenz @laurinlorenz.bsky.social · 14/10/2025Politiker, Journalist:innen, Dissidenten, aber auch Private wurden jahrelange überwacht - ohne deren Wissen. Die #SurveillanceSecrets Recherchen zeigen den bisher wohl größten bekannten Datensatz aus der Überwachungsbranche - mit vielen Verbindungen nach AT... 1/3 youtu.be/U6vIlLueOPUyoutu.beDie Firma, die dich überall finden kannYouTube video by DER STANDARD 26233
Reposted by Gabriel GeigerLighthouse Reports @lighthousereports.com · 15/10/2025🧵What’s the biggest lie the surveillance industry tells? That they only sell to legal clients. That there are red lines. But what do these companies say when they think nobody is watching? We went undercover to find out 712679
Gabriel Geiger @gabrielgeiger.bsky.social · 15/10/2025Muito obrigado! Moro no Brasil :). Eu conheço o caso, e muito interestante. Tenho vontade de pesquisar mais sobre ele. Acho que há muita coisa interessante ali. 120
Reposted by Gabriel GeigerCrofton Black @cr0ft0n.bsky.social · 15/10/2025People often ask what it's like behind the scenes at the world's top spy tech trade fair, ISS World. Turns out it's like this bsky.app/profile/ligh... 21210
Reposted by Gabriel GeigerMother Jones @motherjones.com · 14/10/2025EXCLUSIVE: This tech has been used to track heads of state, journalists, and maybe even you. @lighthousereports.com uncovered a vast surveillance empire whose untraceable tech has been used to target people in more than 100 countries. It has operated under the radar for decades. Until now. 13451239