This investigation exposes the dangerous arrogance of technocrats who prioritize flawed "innovation" over the survival of the poor. It proves that when "data-driven" policies ignore human reality, technology becomes nothing more than a sophisticated tool for systemic oppression.
安装我们的扩展，即时搜索任意视频内容
THE EXPLAINER 5TH MAY 2026本站添加:
On this fifth day of May, we learn of a pact between President William RTO and Tanzania's President Samya Salu.
Just beat them up. Tanzania's President Salu reveals that she and President R agreed on a common approach to dealing with Gen Z and pro-democracy activists.
is a big investment. It takes tanga to tangle. President Ravves controversy on his proposed Tanga oil refinery after Tanzania's President Sulu said she was not aware of the plan.
The school dropout reality. How learners mopped up by government during the senior school transition are still stranded at home.
He developed a cardiac eventually developing a cardiac arrest as a result of a heart problem that he had.
>> Autopsy confirms that Dr. Jobaka died of cardiac arrest as family awaits toxicology results to establish the exact cause of death.
Does SHA work for everyone particularly the lower income earners in the country as promised? Not according to these investigative journalists who have put out a piece error by design and they're taking a look at the algorithm uh that SHA has put together the one the means testing that determines how much each contributor pays. What they have found in their investigation is that lower income earners or the poor are being overcharged by SHA while the rich or highinccome earners are being undercharged. What does this mean? What are the implications of the investigative peace error by design that was done by journalists from Africa and censored? Tonight we speak to them to tell us a little bit more about their investigation and what they found.
tonight on the program to explain to us this Sha algorithm and the flaws in it that they say the government knew about but did nothing to rectify. Joy Kirigia who's an investigative journalist at Africa uncensored. Purity Mukami is a data and investigative journalist. John Alan Nam was also an investigative journalist and co-founder of Africa Uncensored, one of Africa's premier investigative independent media houses.
Tonight we speak to them about SHA and the flaws in the system all on the explainer which starts right now.
[music] Live from Citizen TV, this is the explainer with Ivonne Auana.
>> Jokes is our sign language interpreter.
Let's begin the program tonight with Tanzanian President Samya Saluhu Hassan who has called for cooperation between Tanzania and Kenya in dealing with youth democratic activism in both countries.
Samya said a firm approach from the two governments would be needed to counter the youth and safeguard peace in the region. And as Brenda Wanga now reports, her remarks have attracted criticism from former CJ David Maraga.
The agitation started in Kenya and months later crossed over to neighboring Tanzania.
The region's youth united not just through the internet, but also in their shared demand for democracy and good governance. But as the youth find their voice, the region's leaders are tightening their grip.
Tanzania's [snorts] President Samir Suhu now taking the youth head on. In a speech on Monday, she said activism will not be tolerated.
>> [clears throat] >> East Africa.
She didn't just call out the problem.
She also had a readym made solution.
We have to send on this.
[applause] President Samir's approach not only signaled a hardening stance, she went further to dismiss the universal definition of democracy, suggesting it must conform to local nuance and culture. President Rut however steered clear of the call by his host instead focusing on shared infrastructure in the region as a remedy.
>> In doing so we are also addressing one of the most pressing challenges of our time ensuring that the energy of our young people is harnessed for progress, stability and shared prosperity.
>> In Kenya the Gen Zled movement forced a government retreat on tax hikes but at a high cost of lives. In Tanzania, the scars are more recent. A recent Commission of Inquiry report revealed that over 518 people died in the 2025 election period. The push back on accountability and good governance was already in motion even before Samir's pronouncement. Last year, Kenyan activist Bonifas Muangi and Ugandan Agatha Atuhir allay torture on Tanzanian soil. Even the corridors of justice were not spared. Kenyan politicians and legal minds who traveled to Derelam to stand with opposition leader Tundisu faced harassment and expulsion. Former Kenyan Chief Justice David Maraga did not miss his words, calling the coordinated effort an axis of tyranny. He terms the two presidents meeting and discussions on the matter disturbing and abhorrent, saying that the axis of tyranny the two were constructing threatened the region's democracy. Brenda Wanga, Citizen TV.
>> [gasps and sighs] >> President William Rut says the 2.5 trillion shillings joint oil refinery project set to be built in Tanga Tanzania will not only reduce the cost of fuel in the region but also create thousands of jobs for residents of the East African community. Speaking during his historic address to Tanzania's parliament, President Rallied the MPs to support him as he called for more cooperation between the two countries.
Our political affairs reporter Emanuel to reports.
>> It was the second time a Kenyan president was addressing the Tanzanian parliament 24 years since retired president the late Daniel to teacher Mui gave his historic speech to the Tanzanian legislature.
President William R who addressed the Tanzanian Assembly in Dodma started with an explanation about the deal.
is a big investment.
Your investment is in billions of dollars.
It was informed by events on Monday evening when Tanzania's president Samir Sulu Hassan revealed she was not aware of the 2.58 trillion shilling oil refinery project to be constructed in her country telling the MPs that he came up with idea which he later shared with Uganda's President Yeri Musevveni adding that the deal was later struck during the Africa Web Build summit 2026 held in Nairobi last month.
calculation.
>> The Kenyan head of state further explaining Tanga's strategic position which lies about 2 hours away from the port of Mombasa Uganda pipeline crude pipeline investment President R calling on the MPs to back the ambitious multi-billion project which he says will be a regional game changer.
>> Plastic industry, prochemical indry.
The MPs showered President R with praise, pledging to support the project.
How are we Emmanuel to citizen TV Nairobi >> Back in the country education PS professor Julius Bto has dismissed a Kenya National Examinations Council report indicating that more than 150,000 learners under the competence Y based education system dropped out before completion. However, this denial sharply contrasts with findings from a citizen TV spot check which reveals that a February mop-up exercise by the state department of education did not reflect reality. Learners reportedly placed in various schools were sent home shortly after reporting, casting doubt on claims of a 100% transition rate. As Mary Moi reports, the affected learners still out of school now stand as a stark indictment of the government's commitment to ensuring access to basic education.
Early this year, following a series of reports by Citizen TV highlighting grade 10 learners in Mukuru Kanjangga stranded at home due to lack of school fees. The Ministry of Education issued a nationwide call. In what he described as a mop-up exercise for grade 10 learners, the ministry promised that no child would be left behind and that all affected learners would be placed in schools. According to citizen TV findings, learners were briefly admitted into schools only to be sent back home soon after reporting amid claims from administrators that no government funding had been received and no formal commitment made.
Eric Ma, a grade 10 learner, was offered placement at boys high school in Makueni County during a mop-up exercise led by Dr. William Sugut, the director of secondary education at the Ministry of Education.
He and his mother were assured that all arrangements had been made with Dr. Sug confirming he had spoken to the principal and the school was ready to receive him. However, that assurance later proved inaccurate as a promised placement never materialized.
Eric and many others in Mukuru are now part of a troubling statistic captured in a report by the Kenya National Examinations Council which indicates that more than 150,000 learners from the first CBE cohort dropped out before completing their studies. The PS for basic education has however disowned the report saying it was not released by NEC or any government agency. The report however was released at the NEC headquarters and is written on stationary bearing the neck logo on each page.
>> From where I sit, neck has not released any report. It is an organization which a certain NGO which gave a report and gave some figures which from where I sit we are not sure whether those are the accurate figures that reflect actually the dropout. We do not agree with the figures given the screaming figures are not necessarily accurate. Miriam Moi Citizen TV >> lawyer Paul Mangi and a one-time confidant of the late ODM party leader Ria Odinga made detailed presentations before the ODM party retreat on how Ria Odinga crafted coalitions ahead of any election. Paul Mongi outlined the steps the party should take even before sitting on a table with any political party to talk about pre-election coalitions. Steven Lato with more.
[music] >> The late Rio Dingga was a masterful architect of political coalitions, earning a reputation for his ability to align with former adversaries, stay relevant, and secure positions of power in government.
On a three-day retreat in Mombasa, Ria's party held lengthy deliberations on how their late party leader used to ink coalitions with other parties.
>> Coming to make a presentation today on the issue of coalition building uh because ODM is about to go into its negotiations.
Let Rio Oingga's lawyer Paul Wang bit by bit unpacked how Rio Oinga will start coalition talks.
>> Who can ODM negotiate with and who can it not negotiate with and there is uh pressure from some quarters out outside the party that uh seeks to determine uh who they should talk to and who they should not talk to.
Tender tender.
[cheering] >> Ready. The Orange Party has made it public that it will begin coalition talks WITH UDM.
With factions of the party pulling apart, the late Odinga's legal advisor told the ODM gathering that Odinga will hold talks with his fiercest critics for the sake of power.
>> For Raa, there was no person that was ever beyond redemption. And Ria always gave everybody a chance to redeem themselves. And with coalition talks often proving a complex affair for political parties, ODM is already crafting its coalition path and is yet to present its team of negotiators as Odinga will do. Manga however told the ODM top bras not to look for political angels to negotiate with, but instead push the party's interest at the table.
There are no there are no angels out there and uh you can't go into this negotiation looking for for angels to negotiate with.
>> ODM leaders were told to identify key areas of concern, present precise demands during the talks and accompany them with an implementation matrix then discuss the kind of government to be formed. Mangi saying Rio Dingga never presented moderate or easy demands on any coalition talks table. reform agendas of ODM and that RA stood for are very radical uh who is willing to accept the radical reform proposal that Raa stood for.
>> Wang is saying that before heading to talks with President William R's UDA party ODM must first assure the orange support base that they have the radical ODM on the talk table. If you are going to talk with UDA, the first thing to satisfy yourself is that UDA is ready to accept the radical reform proposal that Raa stood for and uh if they are willing then that is when you go to the next stage.
>> The ODM leadership is back in Nairobi with no tentative talks schedule with UDA or any other party as yet. Stein leto citizen TV >> Nairobi hospital director Dr. Job Obaka died of a cardiac arrest. A post-mortem examination conducted Tuesday had three pathologists in agreement with the family saying it is satisfied. Yet, the defense accuses investigators of sensationalizing the case. Audi Francis takes us through the forensic findings, the suspect's statement, and the exhibits now under government analysis.
>> At first light, the family of Dr. Job Obaka Nairobi hospital's director arrived at Lee funeral home. They came to witness the postmortm examination of a man whose sudden death has shaken the medical world.
Three pathologists John Sen for government, Petua for the suspect and Joseph Nongo for the family all agreeing on the cause of death. And uh we concur with the findings. Uh the cause of death in this case uh he had a cardiac problem and actually he developed a cardiac eventually developing a cardiac arrest as a result of a heart problem that he had.
>> The family says it's satisfied. The government pathologist did his part and we are quite confident about the preliminary report even as we await for the other confirmatory uh uh [clears throat] findings uh when the toxicology uh uh results finally come out in a couple of weeks.
>> But defense council turned fire on investigators.
>> This document has vindicated her.
The sensationalism about her was unnecessary.
She has suffered a lot of public humiliation.
The DCI decided to sensationalize the matter for no reason. From our perspective as the defense, the toxicology is a formality. We expect the DCI to do an honorable thing of terminating the investigation against our client Bris Wangare and releasing her to go home and mourn a dear friend.
>> Samples from vital organs are now with government chemistry.
Betric Wangare told police she cooked a meal for Obaka. He ate, retired to master bedroom and collapsed.
A house search recovered two plates, a spoon, a water glass, assorted drugs, a blanket, bed cover, wet shaw, and a half consumed wine bottle. All exhibits for toxicology.
>> So we await other tests, toxicology, and hisystologology.
>> On May 1st, Obaka metric at Olive Court in Kitangela. Hours later, he was rushed to Nairobi hospital and pronounced dead on arrival.
Burial is set for Friday, May 15th in Musand village, Mumas, Kakamega County.
Od Francis, Citizen TV, Nairobi.
The 12% wage increment announced by President William Rut on Labor Day has triggered a standoff between employers and workers representatives setting the stage for a high stakes battle over who benefits. Court Secretary General Francis Atoulli has pushed back against the Federation of Kenya employers after it indicated the payriseise would only target minimum wage earners. Ati dismisses that position as misleading, insisting the directive was a broad-based increase for all workers.
This as a section of workers led by former deputy president Riadi Gashagwa challenged Awi's leadership accusing him of abandoning workers and calling for an alternative workers organization to rival Kotu. Willie Lucer reports.
Barely a week after President William Root unveiled a 12% general wage increase during Labor Day celebrations, a fears row over its implementation has erupted. A >> 12% increase in general wages.
>> Workers would pin their hopes on fatter pay slips this month now find themselves caught in a crossfire of conflicting interpretations from Kotu and the federation of Kenya employers. minimum wage increase. You >> boss Francis Ati has hit out at Jaclyn Mugo, CEO of Federation of Kenya employers, accusing her of undermining wage increase.
>> And the president this time he announced general wage increase once the gazette notice is issued that is a policy directive to all employers and they know. Ati says FK's position to limit the increase to minimum wage earners is misleading. He's urging Kenyans to disregard the employer's body, insisting the government's directive is clear and applies to all workers.
>> I want to tell Kenyans to ignore the statement, the press statements being released every minute by the Federation of Kenya employer. If Jacqueline has her own government, let her tell that government not to implement 12% general wage increase. At his remarks follow a statement by FKCO Jaclyn Moo on Monday, May 4th, indicating that employers will strictly adhere to legal minimum wage guidelines with salaries above that threshold remaining subject to individual contracts and not automatically applicable to all Kenyans.
forever.
>> But even before FK's position settles, former deputy president Riad Gashagua and a section of workers are already calling for the formation of an alternative workers organization and a shakeup of Cotu. They accuse Francis at abandoning workers and siding with the government at the expense of ordinary employees. It's about time we created another alternative organization to champion the rights of the Kenyan workers. He's at he has abandoned the workers. He is now the lead praise singer of the William R praise and worship team to praise a man who has made the lives of workers hell. Speaking during the national reflections forum and the launch of the labor market and workers welfare report 2026, Gashaga said workers are overburdened by deductions he argues pay sha ns7 housing levy deductions have left many households struggling. go to the ballot and ask do I want to restore the dignity of my pay slip or do I want the oppressor to continue raiding my pay slip and subjecting me to poverty. Pay slips are increasingly show that workers cannot afford mandatory reduction. We will use peaceful action and the ballot to elect leaders who prioritize workers interest. Trade unions must return to defending members.
>> Willie Luc, Citizen TV, Nairobi.
Police in Transmara South are pursuing armed raiders who struck Kauro village in broad daylight, fleeing with more than 130 head of cattle. A community now shaken. Families counting painful losses and residents demanding urgent security intervention as tensions continue to rise along the Angata Barcoy border.
Chrisino with the latest update from Transmar south.
>> Morning breaks over Kauro village but it carries no hope.
Only silence heavy interrupted by grief.
Villagers move from one homestead to another, not in greeting, but in shared pain, trying to piece together what remains after a ruthless raid.
Each compound tells the same story.
Empty livestock sheds, broken livelihoods, and questions with no answers.
At least 25 families are now living in fear after armed raiders said to have been lying in wait in a nearby thicket stormed the village in broad daylight.
[screaming] Gunshots pierced the air. Panic spread and within moments over 130 cattle were driven away towards Tanzania border.
Antony Kirui, a resident, is drained, his breath heavy, having just returned from a desperate but fruitless search.
In his hand, a crude weapon, perhaps more for courage than defense.
Moments earlier, he had stood helpless, staring down at the barrel of a gun as 15 of his cattle were forcibly taken by armed asalants.
For 85year-old Jen Kim Thai, the loss cuts even deeper. Her only source of survival gone.
Fear still lingers in every corner.
Residents recount chilling moments of being cornered powerless as armed men took control.
Locals accusing authorities of failing to protect them, demanding the immediate return of the livestock and action against the attackers.
>> So far, we are following leads from the members of the public who are giving us crucial leads and who are heading somewhere. Those who are involved, we are soon going to get hold of them. And as the search continues, so does the fear. A village left exposed, its people uncertain of what comes next.
Chris Pino, TNO Citizen TV Transara.
Let's take a break on the program tonight, but a reminder of the conversation that is coming up on the explainer tonight. We've invited um investigative journalists from Africa uncensored [music] to explain their piece termed error by design which is an investigation into the algorithm that is driving SHA premiums. Per their investigation they say that uh the SHA algorithm um ensures or has people of lower income being overcharged for their premiums. What does this mean? What effect does that have on universal health coverage? These three journalists will be joining us to talk about that.
Pirim Makami who's a data and investigative journalist, Joy Kirigia who's an investigative journalist at Africa uncensored as is Jean Alanamu also co-founder Africa uncensored. I'm already seeing some of your tweets and questions you can pose to them about their investigation. The hashtag is citizen explainer. We'll be back with them and more news in a moment.
Coming up on the explainer sha and the numbers mystery, we speak to investigative journalists from Africa uncensored on their investigation on the flaws in the SHA algorithm that they say is overcharging the poor.
Sometimes I need a fragrance to keep me calm. Stays fresh in the heat and feels elegant into the night. Lasting longer than expensive perfume. New Town. Love and a calm sunrise. Fresh and [music] sweet elegance. Freshness that inspires confidence.
At Standard Chattered, family businesses, business owners, and corporate executives have entrusted us with over $2.3 [music] billion. Your wealth is guided by certified relationship managers who are supported by insurance and investment specialist. Now is your time for wealth.
Join Standard Chartered today.
>> Imagine a delicate scent so quiet yet so comforting. A mild fragrance [music] that gently drifts you into pure serenity. That's Downey Pure Bliss.
Crafted and infused with natural essential oils. [music] It kept my clothes fresh all day and into the night.
>> You smell like pure bliss.
>> New Downey Pure Bliss. Try it for only 20 shillings.
>> I wish I could have soft and glowing skin. New Geisha black soap with activated charcoal to give you soft and glowing skin.
>> New Geisha black soap and new geisha maringa.
Would you like to work in [music] Dubai or Saudi? M Aira is connecting you with top employers [music] in these and many more countries. We have openings for jobs as taxi drivers, bike riders, cleaners, security guards, construction and many more. We offer full recruitment [music] support from application, placement and guidance on visa processing and onboarding. To apply, visit our website [music] www.m-agiraa.com.
Complete your profile [music] and our systems will automatically march you to the right jobs. Find us on Tik Tok at m -ajira.digital [music] or Facebook at m- ajira or call 0709-488881 [music] for more information. M Ajira, your gateway to credible [music] global opportunities.
Apart from what you say, [music] sour petroleum jelly [music] repairs dry skin like no other. Easy to spread and non-sticky like no other. Great quality [music] product from Pani Oil.
The family of 14-year-old Jackson Tito, the viral boy or the boy who went viral from Dand Pandora, whose plight as caretaker of his siblings, despite being a minor, was highlighted on social media, has finally been relocated to a three-bedroom house in Machakos after years of hardship. The boy who was documented singlehandedly taking care of his four siblings in their shanty house in Dondora's informal settlement on several occasions.
Fanfair greeted the arrival of Jackson Tito and his family as they made their transition from Dand Pandora to Machakos from a life of want to a dignified existence. The fortunes of this family appear to have changed overnight.
change.
The young man who had been left to fend for himself and his siblings in the absence of his mother had been forced to put his education on hold. But that has also changed. His mother has since been reunited with the family and is helping them settle in their new abboard.
Tito's siblings have been enrolled at the nearby St. Mary's boy school. The house was a realization of a pledge made to the young man by President William RTO and is one of the affordable housing units in Machakos County. Brenda Wanga, Citizen TV.
More than a 100 families have been left counting heavy losses after bulldozers flattened high-end homes on a disputed piece of land in Marui area along Nairobi's northern bypass. The demolitions which began Monday saw multi-million shilling houses reduced to rubble under tight security with police officers deployed to oversee the operation. Residents watched helplessly as their homes, some still under construction, were brought down. Citizen TV's Ben Kirui has more.
>> Until Monday, this palatial multi-million shilling residence stood prominently on the contested parcel of land. However, the property had long been marked for demolition after a court found that the land on which it stood had been unlawfully subdivided and sold to unsuspecting buyers by a private company.
Demolition crews moved in with the heavy machinery, tearing down perimeter walls before proceeding to flatten the high valley homes.
By Tuesday, the once admired structures along the northern bypass had been reduced to rubble as devastated home owners watched helplessly.
Do you know police officers were deployed in numbers to oversee the exercise, enforcing a court order that authorized the legally registered owner, Langton Limited, to repossess the land. Some of the affected buildings were still under construction with owners preparing to occupy what they believed were their dream homes.
Despite the shock, there had been prior warnings, including public notices cautioning prospective buyers about the disputed ownership. The property has been at the center of a protracted legal dispute since 2019. Earlier this year, the environment and land court reaffirmed Langon Investments as the legitimate owner and ordered the cancellation of all irregular titles.
Court documents indicate that an entity identified as Miron Limited, allegedly undertook an unlawful subdivision of the land. The court declared the process illegal and directed that all titles and certificates issued from the subdivision revoked, paving the way for repossession by the rightful owner. Ben Kiri, Citizen TV, Nairobi.
Administration police constable Peter Witi, the accused in the murder of Karatina businessman George Gatu has been freed on a 1 million shillings bond and a shortity of a similar amount 107 days after he was arrested. In granting the bond, Ner High Court Judge Kizito Maggare said the defense failed to convince the court of compelling reasons to keep the accused detained. However, his release was tied to strict conditions that include a 1 kilometer restriction to the scene of crime and the victim's home as well as a gag on handling any weaponry until the case is heard and determined. At the same time, the same high court remanded for a further 14 days. Peter and Juguna, the accused in the murder of an 8-year-old minor in Riti village of Madiraa in February pending the hearing of his bail application. Juguna had earlier on plead not guilty to the murder charge leveled against him when he appeared before Lady Justice Moren Odiero.
Coming up on the explainer sha and the numbers mystery, we speak to investigative journalists from Africa uncensored on their investigation on the flaws in the SHA algorithm that they say is overcharging the poor.
Mother's Day is just around the corner.
It's the perfect time to show our mothers how much we appreciate and love them. At House of Leather and Gifts, we have thousands of unbelievable gifts at incredible prices such as leather products, message mugs, message frames, vases, and flowers, wall clocks, gift hampers, [music] kitchen products, and so much more. You can also order online at www.houseoflekenya.co.k and we will deliver to your doorstep.
House of Leather and Gifts, Kenya's number one [music] gift store. Over 50,000 different items at the guaranteed best prices in Kenya. Feel at home more in a slap Bora Group.
>> Aloe body lotion infused with nature's finest ingredients, goat milk.
See the radiant glow and embrace the natural confidence that flows from within from Tropical Brand.
[music] >> [music] [music] [music] >> Visit your nearest NASK local or shop online for more and save big. NAS saves you money. [music] No matter how chaotic your day gets, the new Nice and Lovely [music] Ultra Fresh Dio with zero alcohol keeps you fresh all day with 48 hours sweat and odor protection. Make moves pillar dash.
Try the new Nice and Lovely [music] deodorant today. Nine out of 10 women said they would switch.
Ugali and Sukumawiki. Boil water in a pot and gradually add maze flour while stirring to make thick ugali. Cook while turning until firm and smooth. In a pan, [music] fry onions, add tomatoes and sukuma wiki. Add salt and cook till tender. Serve ugali with sukuma.
This Mother's Month, My Da is showing up for you. Enjoy buy two get [music] one free on select baby diapers and amazing deals on skin care, supplements, [music] and more. Shop now on the My Dawa app or mydawa.com. My dawa, we deliver care.
>> [music] >> What is this?
>> A formal request for divorce.
[music] >> Cuz we wasted so much time.
You know what?
This is the moment I was praying and hoping for Jeff.
>> We were eight people in that country.
>> What do you mean EIGHT PEOPLE?
>> PEOPLE DIDN'T come [music] exams.
[laughter] So I [music] become extremely famous and extremely broke.
This is no no we have to talk. Where are you from? I've been auctioned bro. Fred Mundi hit me the hardest.
>> You going to do it in 2027?
>> Yes.
>> Where?
>> I am Erici.
All right, welcome back to the explainer. Tonight, the conversation we've been telling you about. It's about an investigative piece termed error by design. And it's been done by investigative journalists at Africa and censored and it's all about the SHA algorithm and how it works to calculate the premiums of those who make payments to the social health authority for universal health coverage. We shall talk more about what their investigation found and they are here with me. Let me start by introducing you uh to those who are here with us in studio tonight.
We've got Joy Kirig who's an investigative journalist at Africa uncensored. Next to her is Purity Mukame who's a data and investigative journalist. Ladies, thank you very much for being with us. Also joining us online is John Alan Namu also a face that you have come to know in this country is an investigative journalist as well and co-founder of Africa Uncensored. Ladies, gentlemen, thank you very much for joining us on the program tonight. And by the way, if you've got any questions, please do send them in or your comments. The hashtag is citizenexplainer at Citizen TV Kenya Ivon Aquara or a text a message to the number 242. So let's talk about this investigation um for a moment uh to begin with uh Joy um maybe you can just give us like what the headline of of this is and um you know the sort of basic summary headline of what your investigation found before we get into the details of it. All right. So, uh, thank you for having us. Uh, so this investigation error by design is basically what we've managed to do is is we were able to get into the SHA system based on the information that we got in order to find out why, you know, how it works. So, what we we managed to do is to try and understand how this system calculates uh the premiums for Kenyans.
So that's basically what we did and we wanted to understand because we understand that SHA was uh when it was created was to bring this massive group of people who are in the informal sector. So what we wanted to understand is how is the government able to estimate how much a mamoga is earning how much a border guy is earning so how do they estimate that so that's how we managed to get into the system based on the data that we were given. Uh I know we're going to discuss how we managed to get that data and you know and what we found out was when we managed to do all these things audit the system what you managed to find out is that it overcharges the poorest of this this group of people who are in the informal sector and undercharges the richest group of people in this same informal sector. So basically that is what is our story that that's that's what our story is about. We talked to a number of people from the informal sector. People have been affected by this system where they've been overcharged and they're not able to you know afford uh that their premiums and therefore that trickles down to how uh to to how their struggles in accessing healthcare and we also try to reach out to the government. We also talked to experts some of them who are initially part of uh building this particular system so that we can understand a bit more about what this story is about. So in a nutshell that is what this investigation is is all about.
>> Yeah.
>> So uh purity what what prompts this? So what triggers you into saying hey let's take a look at uh the SHA system means testing all of those things we've been hearing about in the leadup to its roll out. What what prompted this whole investigation to begin?
>> Yeah first of all thank you for having us Ivonne and uh we'll continue from where Jay has left it. In fact, when we say we got inside the system is for lack of a simpler word to say we looked at the formula that calculates the figure that you get assigned to you when you register to the SH system particularly when you are someone who is not uh in formal employment and what prompted this back in 2024 March I was reading um stories that had been shortlisted for this global awards called Sigma awards for data journalism stories. Definitely I mostly do data stories and I came across this >> project called suspicious or suspicion machines and this system um this this project was exposing this unfair and biased risk um scoring algorithm that was being used in Rotterdam to to find welfare cheats like people who want to lie so that they get this to get government support. And what this uh journalist from Lighthouse Reports found is how this system was discriminating some people based on characteristics that they had no control of like ethnicity, gender, uh parenthood and I got very interested that that somewhere much >> two weeks down the line I'm just going around um my everyday work like other investigations that we are working on and Having joined the journalism space, I got accustomed to to being present to current affairs. So, usually instead of listening to music, you'll often find me listening to government to parliament conversations. So I had this live uh session. It was in the afternoon and suddenly this member of parliament from was having this conversation with the then CS and he's like did you just say can you confirm that proxy mean testing is an algorithm?
>> Yeah.
>> And I'm like algorithm. Okay. So I remember that and I'm quickly like could this algorithm be unfair or not fair?
Can we check if it will be fair and all this? So this is how it started >> and he was also talking about um AI and its use because that is what uh you know the government had been talking about that they're using artificial intelligence to be able to do the work of means testing which we'll come and explain uh in a moment. Absolutely. And so that piques your curiosity.
>> Yeah. I even gave an example of this mamai this 87y old window and he was asking can you explain to me how this AI >> or algorithm is going to know the reality of this mama since you're saying this proximing is going to be looking at characteristics of people's houses and and he was like explain to me how this system will pick the realities on the ground yet he knows AI is blind to some realities So that's the day I hand this and I started talking to everybody including John Allan about this idea. I really wanted to understand this algorithm.
>> Okay.
>> And that's how we start thinking. Yeah.
>> So Joy let's now get into this a little bit and it's a little bit technical so we'll we'll try and uh keep it simple.
Um when the government was rolling out SHA and you know as a means of like we're saying to bring uh more people from the informal sector into making contributions they said they were going to go through what was a means testing exercise to determine the earnings of people and therefore their contributions. So you first had to understand >> um this means testing. So from your investigation and your reporting and understanding tell us about how the means testing uh system for SHA uh would work from your understanding. Um so what I understood is that and having spoken to even the experts that we we managed to speak to is that whereas during the you know for the employed it's very easy to determine how much it is that you earn so they can know how much it is that they can charge but for this group of people who are in the informal sector where we have we have the border border guy that I keep going back to they don't have a stable income today they may make 500 shillings per day tomorrow they may make 200 shillings. So there's not unlike the employed where you're expecting a certain amount every month.
It's difficult to determine how much this person earns. So what the government decides to do is to try and bring this particular tool so to speak to use this formula to try and estimate how much it is that these people are earning. And how it does that is while you're filling out the the SHA, when you're registering for SHA, you answer there are a couple of questions that the system gets to ask you. Are you married?
Do you have kids? Uh it gets to ask you what kind of what type of a house do you live in? It gets to ask you what kind of floor is your house made of?
>> And they also get to ask you, do you have a TV? Do you have electricity? All these questions are the are the questions that now this particular system is meant to try and figure out how much you're worth, how much you is worth, how much joy is worth. And then based on that predicted amount that they they the system gives now the government can come in and actually now make the 2.5. It's a 2.5, right? the charge the the amount that they're supposed to now charge it for you to be able to understand how much is your premium, how much you're supposed to pay annually or monthly.
>> Mhm.
>> So that's how I got to understand the proxy means testing system which is basically they're trying to estimate how much it is that you earn. So it's it's like trying to can you afford this?
They're trying the system is trying to understand can you afford this. But now >> you it's blind to some of those realities like she has explained sometimes. Let me give an example of if I have built my mother me and my siblings came together and built our parents a home and we furnished their home. We bought them a fridge. We bought them a a flat screen. We bought them a a smartphone and all those things. when they go out to fill for this uh for to register for SHA the system doesn't understand it is we the children who built this our parents are not the ones who bought that. So those are some of the >> the issues that that we we were trying to understand how would this system really be able to would it be able to really uh cater to the actual realities of the the people. So that's why we're talking about the overcharging cuz the system will not know that I built my mother a house >> right and those are not her earnings.
>> But now once it determines how much you earn >> then that is what is supposed to determine >> how much you contribute. Yes.
>> Right.
>> Yes.
>> Yeah. Purity >> and to earn to her proxy mean testing.
>> It's called proxy because >> uh instead of asking people how much they make >> and proxy mean testing is not really a new tool. It's been in use like since 1990 when uh some world bank scientists really wanted to figure how can we uh estimate incomes of people who we don't have regular records of their income and they figured if we can go and find out how much a household spends and find the goods and services in their household based on the characteristics of this household.
Um we can we can see how much they consume >> based on this observable characteristics and using the consumption uh uh prediction then we can pro you can use that as a proxy for their income and then uh mostly this tool was being and even here in Kenya it's been in use even way in 2010 to to to find poor people who need government support. So it's a good tool to find to to see who doesn't need the support. So it's like a tool that can help you see like gauge. It's like when you go to the market and the seller is trying to see >> how much can this person afford, how much are you worth?
>> Okay.
>> And and of course once that determination is made about your earnings, then they decide to put you in a particular category to pay a certain amount of premiums. Right.
>> Yeah. So what did your investigation then find with this proxy means testing?
Was it accurately determining how much people earn and what did that then mean for the contributions that they make to SHA? So as you say for some time they would put them in category but as we found out during the investigations the investigation they used different methon sometimes put in categories some other time produce an actual number uh but basically for we we onetain and when I say on detain is we trying to reconstruct the model for a lack of a better word I wish I didn't have to use the word model or the algorithm. We reconstructed it because we asked them to give us the everything about the model so that we can check so that we can understand how it's working and test it. M >> so since we were not able to get that we reconstructed based on the information that were in the regulations of 2024 where they gave the formula and they explained the data with which they they trained the formula and we used a section of that data to test. So when we say about this audit, we tested it on the same data that a section of the data that they had trained because what happens is they have this model which they make from the existing surveys so that they can now use the result of that >> to other people who they do not know the the their information. Okay. So what we find uh we have like quite a very detailed methodology that is now published by the lighthouse report.
>> But the most striking finding >> was the extent with which this system or this model was overcharging the poorest and the extent at which it was undercharging the richest among the informal uh uh sector. Mhm.
>> So the model works for some people >> but when at the very extremes um it it really doesn't work well but it's the extent at which let me try and say it in a >> normal like let me not be a statistician. So if I if we take people from the informal sector and arrange them based on their wealth or their income in groups of 10.
>> Yeah.
the the the poorest >> the lowest band >> 80% of them would be overcharged.
They'll be given a figure uh higher than what they can pay >> than what they can actually >> uh than what they actually have.
>> Okay.
>> So they're given a figure of earnings than what they actually earn and therefore given a higher premium to pay.
Yes.
>> All right. joy. Do they are they able to then pay this?
>> Unfortunately for the people that we managed to speak to, it was a struggle >> and I mean what what was a bit um disappointing for me is that >> the government knew this >> before it was launched >> because what we realized is that um and we got uh some some >> So tell me how do you know the government knew about this flaw in the system? Yes, the system uh this particular algorithm or this formula that they use comes with an inbuilt flow. It either overcharges or undercharges. That's actually how it's designed. So when we say error by design, it's actually comes like that. It's it will either overcharge or undercharge.
>> So the government knew this when they they decided to use this particular system. But what is saddening is that there were the government contracted some consultants to be able to review this system. Once they figured out that it actually does this, it comes with those two errors. They brought in this uh consultants to be able to review and give them findings or give them explain to them can we really go with this thing. So the consultants do come back a few days or before the launch and they give their findings and they tell the government >> the government their findings was basically this system is a disaster. It will do this. It will overcharge people.
We we should not be using this particular system. So uh that's that's information we actually got from this uh particular document that we found out that was never public. So we managed to get access to that document and we got to see actually I had written some of the things that we managed to see. It found that this system was a disaster.
They described it as unfair, biased and based on out outofdate data. It also said that this system is highly likely to miscalculate how much people should pay especially the poorest. So the government was again told that this should not work. And when was the government told us by this? Um, >> right before SHA was launched. So this was I think a few weeks before or a few days before SHA was launched.
>> And do you know if the government responded to this organization that did this audit of the system?
>> So we did reach out to the government for a response and we we I mean we wrote them a couple of questions. I think there were about 53 questions for them to answer us so that we're able to understand where they were coming from.
And we also raised this issue that we found this document that has said that they notified you guys or warned you that the system was not going to work but we did not really find out what if the government took in their advice and tried to change the system. So that's why we reached out to the government and tried when this these consultants came back did you try and at least to adjust this particular system >> and what did the government say to you?
>> They've not given us a response yet.
>> All right. I'd like to bring in John Alan Namu now um to talk a little bit more about this. Uh John, so what does this all mean? We're hearing about an algorithm that is uh overcharging um uh those on the lower income scale and undercharging those who are able to pay their fair share. Um this is talk about algorithms, about proxy means testing and about, you know, what people are supposed to pay. But what does this all mean for SHA for universal health coverage and those uh who are watching us tonight?
>> Well, it means that if you follow the intent of the government to deploy this system, despite the fact that they got advice from these consultants, as Joy has just said, then what you're going to get is a a system that's deeply flawed, that's going to be injurious to members of the public, specifically the poorest of the poor in our society. And it'll be one other example in which technology was used supposedly with the intent of helping people but winds up hurting people. And unfortunately, we have a history of this in in our country where we think that technology deployed by itself is the solution when it actually isn't. Technology, like every other tool, follows intent. You if you have a hammer in your hands, you can use it to hammer a nail or you could use it to to bash someone's head in. Unfortunately, the the government seems to tend towards the latter when it comes to these kinds of things. And this isn't just a debate in Kenya. This is a global debate about artificial intelligence. Artificial intelligence, let's remember, at the end of the day, is a tool. It's a powerful tool, but it's a tool nonetheless. And when it's deployed with the intent of, you know, either helping people, then it can be very helpful. It can be significantly helpful. But if you don't follow the intent, then that's where we get problems. And but what would you say to those who say AI is here, technology is here, and the government is employing these tools um to be able to uh be everywhere the government cannot be.
>> I'd say the same thing. Look, artificial intelligence is a tool. At the end of the day, it's a powerful tool, right?
But it is a tool. If you use a tool badly, then you're going to get a bad outcome. despite the power of the two and we see this everywhere across the world where artificial intelligence is being deployed badly. I could go into many examples of the same but you see when when we talk about the government being digital and the government leaning on artificial intelligence I think that the attitude that that kind of posturing is important it's good for us to use the best technology available to ensure that we that we give the best outcomes to you know members of the public wherever they are but if you're being warned that the tool itself is problematic that it is going to injure or it's going to be harmful to members of the public then you need to think about what you're doing. Remember at the end of the day the data that you put into artificial intelligence systems and models right is just as good as the result. If the data is good then the result will likely be good. If the data is bad no matter how much you try and play with it then the output is going to be bad. You know, you know the colloquial say garbage in garbage up, right? And that and this is something that I think the government ought to pay attention to especially now that we are debating what our artificial intelligence policy is going to be and that we're deploying these models across other systems. I think Joyce said it in a story that we're deploying it in agriculture, we're deploying it in education, in some of the most critical sectors of icon. if we do not have a more a richer understanding first of all of the tools that we're using and B if we do not interrogate the intent the intentions with which we're using this tool then we're going to wind up in the situation that we're in now >> joy what does this mean in practical terms because we're talking about a system we're talking about flaws uh in a model you spoke to uh real people um what did this system mean uh for them and how did it affect them in real terms? So in in in real in reality you know this system while we are dealing with numbers in in reality the guys are actually suffering on the ground and we spoken to quite a number of people uh some in the documentary and some that didn't manage to get into the documentary and a case in point was one of the characters that we spoke to in the story where she was she got a higher premium than she could afford and at the time she was pregnant and she was expecting a baby. So at that particular point, she really needed to figure out where am I going to get this money because the minute now uh you know it's my it's time for me to have my baby. I need to have at least something you know this insurance that can cover all these things. So this this lady that we spoke to uh the husband actually she's jobless. She doesn't work but the husband works in the informal sector. So that means their income is it's really something that it's not um >> guaranteed every day.
>> Guaranteed every day. So while this thing while they were overcharged, they did their best to try and you know deny themselves anything else so that they can be able to save up as much as they can in order to be able to cater for themselves when their baby was coming.
Uh but now >> they able to pay the premium because you said uh in the piece um how much were they asked to pay >> around >> 650. Yeah.
>> So so uh >> actually when she says over a change it's not a matter of perception or what this person feels >> right. It's a case of when they were asked to pay 550 >> and we trying to estimate like how much this uh the husband makes from the mjango job >> and like on estimate you would make 12,000 and when we calculate 12,000 in per month >> and when we calculate 550 >> of the rate of the percentage is 4.58% And remember the law says they should pay 2.75.
So they already overcharged. And we talked to many others. In fact, we talked to one uh household worker like daybag >> who who pays who get like the higher level the higher figure for house helps 15,000 because she works in this uh affluent area.
>> But she was asked to pay 1,300 >> every month. uh per month in total would be a different figure and this would be 8.675%.
>> That's what we mean that they are overcharged and that it's not like they feel like I cannot pay. It's >> actually it's higher than the >> it's actually higher than what they should pay by the law. I see. So then what did they do like uh the lady um you know who was expecting were they able to make the payments the sharp payments? I mean they they did manage to pay for the yearly but at least they had enough to cater for a few of the the needs that they needed while they went to the hospital.
>> They paid half because they started in the middle of the year. So they were able to pay for like 5 months by the time they they got the baby after that >> so so after that as for this year when we were talking to her she had not paid the premium for this year. for this year.
>> She hadn't paid at all for this year, but for last year, she was able to pay half.
>> I see.
>> Yes.
>> And when she's not able to pay for this, >> what does that mean? What's what's the reality? That means what?
>> She was disappointed because I think at the time when we were interviewing her is when she learned there's an there's actually an appeals mechanism. She did not know that existed. when when she was doing the registration, she had hoped that she'd have been told if you feel that this thing you've been overcharged, you can actually appeal for this process in order to >> and you did the appeals with her.
>> Yes, we actually did an an appeal live on camera for her to see, but the appeal was denied. Okay. Yes. It didn't work out for >> But this means what in terms of accessing health care? I mean if she's not paid the premium uh she's not started paying for this year then it means it's difficult for her because when she goes to the hospital we cannot tell if she's going to access the the you know health care the way she's meant to because she's not paid. So that's where people are are are struggling where they're like if I'm not able to pay why am I being punished and it's not my problem that I cannot afford to do all these things yet you've been overcharged. So I I feel like that's where most of Kenyans are are struggling, >> right? Um I I want us to take some questions uh that are coming in um for us in a moment. But you know this investigation and the people you spoke to is is not new. People had been complaining online as well, right? About the premiums and how some people were being given premiums that were higher than than they could. Um so beyond all the technicalities, what what do we do with this information? um you know now we have this story and people say okay well maybe this is this is technical all systems have flaws joy um what would you say to that >> well what I think it's important like uh when we started this I did not understand all this uh language that fit had been handling before algorithm and all that but what I realized is that most of the people like I was giving an example of a border two border border guys >> what I realized is most of the People are confused why we would be working at the same working on the same job like if it's mjango or like as a border guy but when we go to register for this I get a higher premium and you get a lower premium so what we learned by auditing the system is that these small things these questions that are being asked are used to determine it one of the shocking things that I learned that even your level of education can be used to either you know over or uh to have you charged higher or lower electricity If you have electricity in your house can also be a factor that leads to you either being overcharged or undercharged. But you see those are some of the things we're talking about that if if I got electricity from the previous government through the rural uh electrification project that the government was doing. I did not connect myself to this house.
It's the government that brought this thing. But when I go to register the system sees that I actually have electricity in my house. But it's not me who brought this. So the system uses that as far as you have like something electricity. If you have a smartphone, if you have all those realities that I gave about me, you know, building the house for my mother. So that for me was like if I could answer that question where people are confused about why purity and I are doing the same job but one is overcharged and undercharged. One of the key things that people really needed to understand is some of these factors like your roof, what type of roof can be used to actually estimate whether you're worth higher or lower.
the floor of your house can be used to to to estimate that. So for me that was really important >> which might not be a bad thing except to understand how you got uh those things whether it was through your own earnings and therefore if you're able to afford to pay more.
>> But if you cannot afford to pay more >> then that means >> you pay it or you don't. You don't likely or you struggle. And if you don't make the payment what does that then mean?
>> What does that mean for you when you go for healthcare? Does that mean you're not qualified? Does that mean you can't you can't enter hospital and get treatment? So, yeah.
>> Okay. All right. Let's take some uh questions. Uh now, uh we have this one from um uh Raz Bashir. What did your investigation uncover that shocked you the most? Were there any attempts to block or interfere with your findings?
Uh these are three questions. And what gaps still exist that need further investigation? Um so, we'll we'll take them one by one. What shocked you the most? Purity, we can start with you just very quickly.
Um what shocked me the most is like the extent the percentage like 80% of the most poor people in the society would suffer and on the on the like way forward what you're asking Joy I ask Florence so now that you've appealed and you've not succeeded so now what she saying she just hopes that her her child and her husband don't get sick.
>> Um but what um I was very disappointed in is how this whole the algorithm bit which I think is the foundation of this whole program >> is how all this information was kept away from us and when we tried to ask for this information how difficult it was for us to get this information >> very quickly um did you how long did it take or what was the process of getting um you know access to the information about uh the algorithm? So the lead through that we were able to get which was like the indicators and the form in which they trained the model >> um took we started sending them emails back in May.
>> Mhm. We got this May May 2025 >> and we got the first response after she she had a lot of back and forth with ombudsman and sh that was around November is when we got some response.
Okay.
>> And it wasn't complete >> right. So you had to go through the office of the ombbudsman. So you first wrote to the government asking for information when they time lapse they didn't >> after how long?
>> Uh so we supposed to wait for 21 days to give them time. So when they don't respond, you can file a complaint with the onboardsman. So we did that and started the process. Then they kept pushing us. They were like, we don't have to go to this particular office.
They they So we wrote to SHA, we wrote to the ministry of health. We also wrote to our digital health authority because we at the time we could not tell who has this information that we need. So we ended up with SHA.
>> So it took you how many months?
>> A couple of months. It took a couple of months.
>> From May to November. Yes.
>> Okay. Um I just want to bring in John.
Yes. Yes.
>> Yeah. I think one other one other response we still haven't got is from the ministry itself to date right we we've published the documentary and the number of months that have that joy and and purity took just to get the the the responses that they got shows you just how much information is still out there I think to answer the third question that that your viewer was asking >> and that to date the government still has not responded to very very wellought out very very comprehensive lists of questions about this system. We're not doing this and and I think this is this is habitual behavior that that we need to start to discuss a bit more. We're not doing this because we have some sort of acts to grind against the government.
This is something that's being done in the public interest and it's important for members of the public to know that the systems that they are paying for with their taxes, with their hardearned money are actually building the thing that the government says it's building.
So what I' I'd hope for is perhaps after this interview that the Ministry of Health led by the cabinet secretary or whoever he deems important enough to respond to us can give comprehensive responses to the questions that we're asking and from there we can have better ground to understand whether there's anything further that we need to investigate because right now there are number of gaps including where the payments themselves go. You know, these are these are very serious matters that I think people ought to be very aware of. I mean, healthcare is part of the foundational pillars of this government's manifesto.
>> It's important that they treat questions that have been asked about healthcare with the seriousness that it deserves.
>> What was it that surprised you the most through this investigation?
>> I think the thing that surprised me the most was that we are still doing this, right? And as I'd mentioned at the beginning of my answers, this is not the first time that we have used technology as this sort of blanket to say that things are going to be good and then further down the line try and selfcorrect. Remember in 2014 an investigation that Joy and Purity were also involved in. There was a man manage equipment releasing scheme that was launched and it was launched ahead of there being consultations with with the county government. If you're looking at election software, right, we know what the problems are with that. If we're looking at um at CBC, etc., and the and the level of cons consultation that takes place between government and stakeholders in the public. This is why we wind up with the problems that we wind up with because a a solution that is supposed to be a one-sizefits all is used and b the consultation is minimal or shingle panda at best and and c this leaves the public with a burden of then catering for health care that they shouldn't have to cater for because they've already paid right as as was said in the documentary this was supposed to get rid of haram days. Are we still paying Harambe today? This is one of the primary questions we need to ask ourselves for in terms of healthcare. And I think that's what surprised me. It surprised me that we are still going down this this same path and yet um not for want of technology, not for want of expertise because there's a lot of expertise inside government that we'd still proceed down this line. Did it surprise you to know that the government knew about this flaw in the system prior to the launch?
>> Um, you know, I I I I want to be kind to to the government and say that perhaps they they didn't know as much as they should have, but but unfortunately I've been in this profession too long to to learn, you know, g give that kind of kindness to to this to this government.
I mean, it doesn't surprise me or any other journalist that has attempted to get answers from the government about this or that has found out that the government indeed knows about something that could be harmful but chooses to deploy it anyway.
>> I I know that there are many Yeah, go ahead.
>> Sorry. Do you think they should have told the public about that report um that was provided to them regarding the flaws in the >> I think it in in due course I think they should have what they should have done is that you know press press pause on the implementation of of um of this system figured out if there was another way to do it then come back and and tell the public look we have you know the with the best of intentions um deployed this system that is using an algorithm that we have found then to be deficient in terms of um you know being able to estimate the income that that that you have. Let us go back to the drawing board. In the meantime, we have albeit an inefficient system, we have a system that is is is largely working. We can continue with that up until the point that we are certain that we can deploy these kinds of things. But you see, when when when um when policy is hinged on political promises and those political promises take precedent over the actual realities on the ground, then you wind up with a cake that's halfbaked.
>> Let's uh yeah, >> I just also wanted to say that they didn't even need the that consultant report. Even I'm sure the consultants who from the word go supported them any statistician knows any mod no model is 100% accurate. So it isn't a surprise that it's not 100% accurate. It's they I'm sure any statistician wouldn't communicate this is the model and this is the amount of error and this is how the error will be behaving and even in this formula that they published the last thing and there is an error. So that is why that part wasn't very surprising for me. What was very surprising is how they were like building the the plane while it's flying because you would see some people filled 17 questions, some other people filled 25 questions like how things kept changing >> and how they kept doing this for such a critical service. Okay. All right.
That's that's that's an interesting observation. Let's take a look at this view from uh Shadi Opio. Shady Opia SH SH SH SH SH SH SH SH SH SH SH system is discriminative um I can't access even with the OTP that's the onetime password two I don't understand I am unemployed and rated at a premium of 12,000 shillings same as under secretary on a review appeal application they rejected it I am now challenged to access and the intolerance of quote unquote system is impuned to our financial ability uh so here is someone who's saying they're unemployed and being asked to pay a premium of 12,000 shillings I believe a month. Let's take a look at some more feedback. This one is from uh J67.
I kind of feel that the government deployed the SHA AI model without doing enough training or tests with real data that is a good representation of the different demographics that we have in the country. Um and then this is from Leishon Laban says watching from Norway.
I appreciate uh the interview on the SHA algorithm flaws. As an investigative journalist into AI and data, I agree with my mentor Namu that AI in government can enable fraud if misused, but health systems must stay honest. And we have another one here from Nixon Duggier. Uh from your investigations, one, how easy is it to lie to the SHA system to conceal earnings? And two, can the findings of your investigations be useful in a court of law to defend the overcharged?
Who wants to take this between the ladies? Namu, I don't know if you're able to hear that uh last uh question.
>> I think in terms of in terms of hiding um [laughter] your earnings from government, I I don't think that that I'd want to answer the question because then, you know, you'd be playing some of the same games that that criminals do. But in terms of how the proxy means testing, you know, data collection um um process works, I don't think that you can because some of these things are done in person. Joy and and Purity who are in the field would be best placed to answer. Um the second the second um you know question is is an interesting one. I think that what Joy and Purity and Gabriel from Lighthouse Reports and our partners at the Guardian have uncovered is sufficient to then form the basis for further investigation that could lead to um you know to to a court case. Whether it's going to be used for the overcharged or the undercharged, I can't say. Um but I think ultimately what we've we've endeavored to do and and I think this is this is one of the markers of the kind of investigations that we do. Not only do we do the investigations but we also show our working as we used to be told in primary school. So if you want to find out more about um what you know what methodology was used um we'll be publishing that on our website. It's already published on uh the Lighthouse Reports website. Um and as well as some of the code that Purity and and Gabriel used to be able to decipher what was going on and rebuild uh the model which is published on GitHub. So you can find out what our investigation was like then I think anybody who's willing to take this further I think um that would be the place to start.
>> All right, let me uh take some more feedback and then we can uh close uh the conversation. Now, let's take a look at uh Hez Washira. Uh so, they're not technicians. Their investigation is based on assumption.
You know, you're going to get this, right? So, [laughter] it's not about >> uh let's let's uh let the ladies talk about this. Yeah.
>> Yeah. I think what I should have said, even though I'm a data journalist, I'm actually a statistician with a masters in applied statistics.
>> Thank you. and Gabriel Gabriel Gerger who we worked with has done similar audits in many countries and in addition to that we had our methodology reviewed by four professors from four different universities. So we are sure of what we are talking about and that's why we publish the the methodology.
>> All right, let's take a look at another uh tweet here. Uh this is Rachel Tabitha. Do they know how much the system cost? uh did they uh do a comparative analysis from other countries using such systems to know their costing?
>> Uh so what we found out is this algorithm or or what people who contribute into it told us it's not part of the contract. We also got to read the contract and the contract doesn't barely mentioned there w mean testing. So from my understanding from the different uh sources we talked to the that contract was to cover very different things uh as from the mean testing tool but we also asked the government about the cost for this and they we haven't gotten the answers. So uh those who contributed to development of the model saying they contributed in form of other funding like how the US would fund so they would >> uh not contribute in monetary terms so we do not know how much it's cost maybe in form of workshops and other discussions we wouldn't know right >> could I just answer on the assumption about whether we are using assumptions we're not using assumptions because the information you used was given to us by the government. What we've used is government data. So we've not we've not come up with any information that is ours. We sent this foyer request and the government was pushed to give us this information and what they gave us is what we worked with. So what what our investigation is actually based on the the government data that they actually gave us.
>> All right. I want us to close this now.
Uh Purity, your uh final thoughts before we we close with some more feedback. Um, so we have this information. Um, what next?
What next? And what I hoped when I was bothering those people who had done this before me to really support us to do this, I hoped that we would give Kenyans capacity or some form of tool or confidence to hold their government to account to know what to ask for when all these systems are thrown at us. It's not just a mention of a system. We should know which formula, which training data set, how are you collecting the data, what's the error rate for your for the for the model. I just wanted ordinary Kenyans who are not statisticians like me and others would know it's simple and that's why I collaborated with Joy who is good at breaking this and this is why you will notice in the documentary we really trying to use simple words because this is something Kenyans can understand and just the same way they hold government to account for corruption and other wrongdoing. It is their right and duty and responsibility to ask government similar questions when they have these systems in other sectors >> including in in health. So your final words?
>> Uh my final words is having spoken to so many people Kenyans are not unwilling to pay. It is the fairness of this thing.
If I am overcharged like most of the people would be say if I was actually charged 300 I would afford it but I'm being charged 650. It's a struggle for me. So actually this whole idea of bringing in the informal sector to to actually contribute to our our you know insurance the the insurance cover it's actually a good idea but how are we going about it? It's not that Kenyans are not unwilling they're not willing to to pay but how are we handling this whole situation? So I would hope that the government would be on would have been honest with this entire thing. uh but we are where we are and one of the questions at the end what do we do next is what the documentary ends at so what do we do next now that we know all these things we've seen the complaints we've seen Kenyans are unhappy we've seen you know when you go to the hospitals there's so many issues that are happening there so what do we do next now that we know that >> so that's what I would say for the government what do they do next yeah >> John closing thoughts >> I think yeah joy took the words right out of my mouth in terms of saying that um you know Kenyans are unwilling to pay. It's it's really just a question of fairness. I think what I'm going to close with is the conversation about artificial intelligence. I think Kenyans ought to remember that they shouldn't surrender their rights to question because the technology seems complicated. The questions that you ask of the the technology that is being deployed in your country are important because that technology will likely hinge on some rights that you have that will either be expanded or in this case, you know, contracted or taken away. At the end of the day, the conversations that we hear about artificial intelligence coming to replace you, the kind of fear-mongering that we hear about artificial intelligence or or saying that it's a solution for everything is neither this neither here nor there. At the end of the day, it is us who have the agency as a country to define how we use our artificial intelligence to tools that we that we have at our disposal and they must always be used for the public benefit.
If that falls if that's something that falls short, if you feel that a tool is not being used for the public benefit, it is your right to question. Do not surrender your right to question technology just because it seems complicated. Alinda Ivonne and thank you very much for having us.
>> Thank you very much indeed for making time for us. John Alan Namu um joining us uh from San Francisco and with me here Joy Kurika. Thank [music] you very much in purity makami. Ladies, gentlemen, thank you uh for um your piece error by design. Let's take a look at one last uh feedback before we close this discussion. This one is from uh Nashan Ado who says to be reliable algorithm based models must be tested rigorously and for a long time with reviews and contextspecific [music] data. Hurried implementation of imported duplicates is an injustice to citizens.
Same case with the failing MIT for banning hel applicants. Um indeed speaking to what you were saying about how this is this means testing is being applied across various [music] sectors um of uh the economy. Ladies, gentlemen, thank you very much [music] uh for being here with us. That piece era by design on SHA and that model to determine premiums that are paid by you. The Kenyon is out now. We thank you for your time and for the work that you put into that. Let's take a break here on [music] the explainer. When we return, we've got the today's business and sports.
Automotive firm car in general rebounds with a 2.4 4 billion shillings net profit up five times from the previous year driven by growth in the motorbike sector.
Not all serums are as fast at reducing dark marks. [music] Choose new Garnier even and bright vitamin C plus booster serum. Combining vitamin C with activated nyinomide and salicylic acid to create vitamin C plus 100 times stronger than [music] classical vitamin C. Clinically proven minus 66% dark marks. Even brighter [music] skin from day six.
New Even and Bright Vitamin C plus Booster Serum [music] by Garnier Naturally.
The county [music] government of Kiamu under his excellency Governor Kimani Matangi notifies all developers with unapproved buildings to regularize their plans as required by law. Submit applications [music] to your subcount physical planner by 31st May. Approval timelines are 7 days for single dwellings, [music] 3 weeks for commercial buildings. Do not pay individuals. Use official bounty journals only. [music] Our bodies get serums and washes, but underarms only get deodorant. That's why our new breakthrough formula has pure hyaluron and vitamin C to deeply care for your underarms for soft [music] and even skin tone in 2 weeks with 72-hour sweat and odor protection. Try the new Nivea Derma Control range because your underarms deserve more. Nivea is for skin.
Faith Kipon is definitely looking strong for her next race.
Hi, I'm Fate Kagon, triple Olympic champion, world champion 210.
This Mother's Month, My Da is [music] showing up for you. Enjoy buy two get one free on select baby diapers and amazing deals on skin care supplements and more. [music] Shop now on the My Dawa app or mydawa.com. My dawa. We deliver care.
You need a bigger space [music] to grow your business.
You need to be ready for any season.
You need faster checkouts for these busy days.
You need more equipment to meet demand.
You need to make sure everyone gets paid on time. In business, you need advice, solutions, and sector expertise when it matters.
>> Your new office looks great. What's next?
>> We get it with Absa Business Banking. Scan this QR code or visit an ABSA branch to speak to a business relationship manager.
[music] The hottest snackberry lunch layers of peanuts, crisp rice, wafer, caramel coated [music] chocolate lunches worth 1.2 million Kenya shillings. From gaming consoles to [music] smartphones hampers what's up high 0706152430 to enter. Gadbury lunch bar f you [music] to do much much more. Decency supply.
>> Kenya, are you ready?
>> Apostle Peter Manuru of Jesus teaching ministry Nairobi invites you this Friday for an overnight kesher.
>> I PROPHESY >> PROPHESY >> YOU WILL GO FAR IN JESUS NAME.
>> AMEN. WE are located along Mananja road next to Harambe Sako in Don Nairobi.
>> Come and receive your blessings and miracles.
[music] Juicy [music] juicy juicy.
>> [music] >> Stand a chance to win playful prizes.
See details in store. Te's apply.
>> Disaster can [music] strike at any time and without warning.
came on a claim from so we will get back to you in three working days but this is where PCF comes in policy holders compensation fund PCF [music] insurance policy holders claimments >> [music] >> 0794582700 PCF.
She raised you, guided you, loved you through everything.
[music] This Mother's Day, celebrate her at Mothers and Melodies Chapter 2, a beautiful outdoor concert in Tiggoni with [music] live performances from your favorite artists in a warm familyfriendly setting. On Sunday 10th May, bring the whole family, [music] Mom, Shaw, Shaw, and the toys for a day of music, connection, and unforgettable memories featuring an amazing artist lineup. Mothers and Melodies celebrate mom beautifully. Proudly sponsored by Inoro FM, Inoro TV, Citizen TV, and Citizen Digital.
Apart from what you say, [music] sour petroleum jelly repairs dry skin [music] like no other. Easy to spread and non-sticky like no other. Great [music] quality product from Pani Oil.
All right, let's talk business now.
Automotive firm car in general has reported a sharp rebound in its 2025 earnings with net profit jumping nearly five-fold to 2.4 billion shillings. The growth is driven by rebound in borda borda sales, stronger regional performance and diversification across its business. The company however says the high cost of doing business in Kenya remains a concern. Here now is Jasmine Wboy with the details.
After two difficult years, Karen general is now back on an upward path. The company's gross revenue rose to 25 billion shillings, up from 21 billion shillings a year before. Net profit jumped to 2.4 billion shillings, a sharp rise from just over 500 million shillings in 2024. A key driver of this turnaround has been the recovery of the motorbike market where marketwide monthly sales shot to about 8,000 units up from 4,600 units last year.
>> Right? So that's quite a big jump if you think about it. It had dropped from 20,000 to 4,600 when our results were really bad. Right? and then it's gone up to 8,000 per month in 2025.
>> The company is also increasingly leaning on regional markets with Uganda in Tanzania now contributing more than half of the total group sales, helping cushion the business against local pressures. Even so, the CEO has flagged rising cost of doing business in Kenya as a growing concern, noting that operating expenses continue to weigh on companies despite the recovery. We wanted to set up a tire factory at one point in Kenya, but the reality is it's much cheaper to set up in Uganda. So, I'm just using an analogy. There are three or four tire factories in Uganda.
Today, we import batteries from UK.
>> From Uganda, right? Why are those factories not in Kenya? Must be because they're more expensive, right? Why not in Tanzania where you have a ship, you have a sea, you have a port. So I'd say that when I look at it, I'd say out of all of them, Uganda is probably the least expensive. I'd say Kenya and Tanzania are on a par.
>> He says the firm's strategy to diversify both its products and markets has been critical in maintaining profitability and staying competitive and in light of the performance, the board has recommended a final dividend of three shillings and 12 cents per share alongside an interim payout made earlier in the year. The company now says its focus will be on efficiency, cost control and growing market share. Now the survival of most businesses in this country is quly pegged on diversification of both product and markets. And this is happening at a time when scaling of operations regardless of how expensive it is is both critical to profits and survival in the market.
Jasmine Wambooi Citizen TV Nairobi.
>> The Intergovernmental Authority on Development EGAD has launched the regional youth coalition on climate and climate resilient agri food system. An initiative aimed at bringing together youth from all member states working on climate related initiatives to enhance agriculture in the region. The initiatives intend to address the funding gaps that key initiatives face while offering the youth business development skills as mentorship opportunities. The coalition is set to embark on a fundraising mission. This will be in addition to the $1.1 million or about 142 million shillings issued by the Alliance for a Green Revolution in Africa Agra which has so far dispersed half of the amounts based on the achieved milestones. The coalition will also push for the inclusion of the youth perspective in the region's policies to ensure that the region which boasts of over 60% of youthful population reaps for the youthful demographic >> funding gap is is is really big. You've seen the youth in the room. They have very great ideas. They have uh really impactful work that they are doing on the ground. So the funding uh gap or need is is quite huge. But all we can do is make efforts going forward, leverage the funding that we have from Agra and see if we can get more resources from other partners to continue supporting the youth.
>> We are particularly encouraged that this initiative goes beyond dialogue. It is anchored on mapping youth actors, building their capacity, connecting them to mentorship and creating pathways for policy influence and enterprise growth.
This is the kind of structured and intentional engagement that young people across our region, across our continent have consistently called for. Kenya's diaspora is increasingly shifting from sending money home for upkeep to financing startups, opening up a new frontier for business growth in the country. A new push led by the Kenya Diaspora Alliance is now linking Kenyans abroad with local entrepreneurs in need of capital. At a business conference in Nairobi, investors and founders explored how diaspora remittances can move beyond consumption into structured investments in sectors like technology, agri business and manufacturing. Partners say tapping diaspora capital could unlock jobs and scale up startups especially as access to traditional financing remains limited.
>> Yes.
So you were talking about diaspora remittances, inflows and investments.
The African diaspora pump into the continent a hooping hundred billion dollars every year. For Kenya it's about 700 billion Kenya shillings. So that's 5.2 billion dollar per year. I dare add that often we have a target. What is it that we can do with government at both levels as well as with diasporas and corporates to double this? Move it to one trillion Kenya shillings by the year 2030. And truthfully, what we found throughout history is that when we don't support ourselves, other cultures, races, and peoples, with all due respect, don't have our people's best interest at heart. Often times >> there are a few who do, but by and large what we see is exploitation, unfair terms and deals, so forth and so on.
[music] [music] Football Kenya Federation Vice President Macdonald Maria and eight national executive committee members have lifted the lid on a 43 million insurance scandal and how the money meant for Chan 2024 insurance agreement was paid to a non-registered insurance brokerage firm riskwell insurance that was not licensed in the country and not listed by the insurance regulatory authority. The company was also only incorporated by the registar of companies 40 days to the start of the African Nations Championship Chan tournament that was co-hosted by Uganda, Tanzania and Kenya.
According to the nine FKF NEC members, the insurance companies set exposed players to lack of cover during the period of the tournament. Melita Oletenus reports. You have uh >> embattled football Kenya Federation President Hussein Muhammad faced questions from the National Assembly Sports and Culture Committee. Over 43 million Kenya shillings insurance payment allegedly approved under his watch but later found to be invalid and that the insurance firm had only been registered 40 days to the Chan tournament 2024.
The nominated neck member Abdullah Yu was accompanied by an individual unknown to the federation. The nominated neck member proposed an insurance brokerage company Riskwell Insurance Brokers Limited. It's unfortunate that as a federation we gave C a fake certificate of issuance they from takul after all this >> despite an expose from vice president Maria and eight neck members raising red flags over financial mismanagement placing FKF president Hussein Muhammad and nominated neck member Abdullah Yusf at the center of the scandal Hussein had earlier denied any involvement.
>> I don't know about this company. I don't know about its registration. I don't know who the owners are.
>> But his denial clashes with evidence tabled before the committee. Payment receipts presented and said to be from Echo Bank linked to the president's email show a transfer of 328,735 about 43 million Kenya shillings made on 4th August 2025. The monies were paid to risk well insurance brokers through the first community bank limited.
>> 43 million was paid as insurance premium to a broker who was not registered broker and consequently it meant that there was no insurance. CF is not in Kenya and for CV to operate it is through this federation.
So there must have been consultation and CV must have consulted you because K cannot come from Egypt to come and know the insurance companies.
>> As an individual I get particularly scared when these things happen because you can imagine as the only person of the alternate gender I will be alone in Langata. At least there'll be 14 in industrial. [laughter] >> The Sports and Culture Committee of the National Assembly now suspects a deeper cover up within FKF, warning that the funds may have been mismanaged despite being under CF oversight. President Hussein Muhammad finds himself increasingly isolated under fire from both his deputy Macdonald Maria and neck members.
Mita Oing Citizen TV.
>> Nairobi United take on Tuscar FC on Wednesday afternoon in one of four Kenya Premier League matches at the Ulzi Sports Complex kickoff set for 300 p.m.
The match was initially scheduled for Naya National Stadium, but the venue was changed to the Ulzi complex on Tuesday afternoon. Nibbos will throw everything at Tuscar FC knowing that a win with will move them up from sixth to fourth on the log on goal difference if police fail to win against Moranga Seal. Naibos were eliminated from the FKF Cup by KCB after losing 1-0il in the quarterfinals.
Their league form has been impressive since April, winning three matches and drawing two with no loss. Narobi United won the first leg encounter 1-0il thanks to a first half goal by Kevin Otiende.
The other matches on Wednesday afternoon will be Bandari against poster Rangers in Mombasa. Police taking on Moranga Seil and APS Bomet facing KCB.
>> We are delighted to see our players coming back from injuries only that likes of Cifa who is a bit injured.
Samimba also has a bit of a knock. But all the players are 100% fully recovered. You saw Olla came in uh the last two matches is recovering well. Um Zeales has recovered. So tomorrow is a basically a very good uh good match.
Finishing finishing ninth finish third asani. So finishing up there motivation considering league this season. So finishing decent position to motivate going into the next season.
Yeah.
相关推荐
Agentforce NOW AMA: Build with React and Salesforce Multi-Framework
SalesforceDevs
490 views•2026-05-28
How agent o11y differs from traditional o11y — Phil Hetzel, Braintrust
aiDotEngineer
450 views•2026-05-28
Re: 🗣️📍theprophedu📍2026 GST 103 CLASS (E-EXAM REVISION)
theprophedu
636 views•2026-06-04
WEB TECHNOLOGIES UNIT-2 | Degree 4th sem BCOM Computers web technologies unit-2 full explanation💯✅
LearnwithSahera
1K views•2026-05-29
More tests are always better? How to use AI to identify tests that bring little value
Alliance4Qualification
335 views•2026-05-29
Search Algorithms Explained in 60 Seconds! 🤖💨
samarthtuliofficial
218 views•2026-06-01
People of Game of Thrones using JavaScript DOM
AltCampus
296 views•2026-05-30
Instagram accounts got PWNed
EricParker
13K views•2026-06-03
