Kizza Besigye, who faces treason charges, is reported to be ‘unconscious’ in hospital after collapsing in court.
Detained Ugandan opposition leader Kizza Besigye has been hospitalised in an unresponsive state after he collapsed during a court hearing on Wednesday, according to his wife.
… %!s()
Besigye, who has been in prison since late 2024 on treason charges, is now “unconscious, unable to speak, and unresponsive”, his wife, Winnie Byanyima, said in a post on X on Thursday.
Recommended Stories
list of 3 items
list 1 of 3Uganda court denies bail to opposition leader in treason case
list 2 of 3Uganda opposition leader Bobi Wine cleared to run against Museveni in 2026
list 3 of 3Uganda cuts internet days before presidential election
end of list
He has been admitted to the intensive care unit at Mulago National Referral Hospital in the capital, Kampala, added Winnie Byanyima, who also heads the United Nations AIDS programme UNAIDS.
Byanyima has called for her husband to be transferred out of Mulago hospital, a public facility, and into a private facility where he can be treated by his personal physician.
“[Besigye’s] family is saying he needs special treatment and they do not trust the government hospital,” reported Al Jazeera’s Catherine Soi from Nairobi in neighbouring Kenya.
“There’s a lot of frustration. The court has not granted that permission for him to be transferred.”
‘Not responding to anything’
Besigye, 70, was seen falling into the dock during his Wednesday courtroom session while protesting against his trial without lawyers of his own choosing, after authorities also detained and charged his main lawyer, Erias Lukwago.
“Before he collapsed, he cried out that he was being injured,” said Byanyima.
Ingrid Turinawe, a close confidant of Besigye, told The Associated Press that she and others were not allowed to see Besigye at Mulago hospital.
“He is in the ICU, and he is not responding to anything,” she said, adding that Besigye’s personal physician was able to see him several hours after he collapsed.
Besigye, a former ally-turned-critic of longtime President Yoweri Museveni, has been in custody since November 2024. He was jailed together with his aide Obeid Lutale in Kenya and both were repatriated to Uganda where they were subsequently charged with treason.
Besigye’s lawyers, supporters and rights activists say that the charges are politically motivated and that his prolonged detention is part of an ongoing crackdown on opponents by Museveni.
Both Museveni and his son, military chief Muhoozi Kainerugaba, have already weighed in against Besigye.
Kainerugaba, alleging that Besigye plotted to kill his father, has previously described the opposition figure as “a dead man walking”. And Museveni himself has said Besigye must answer for “the very serious offences he is alleged to have been planning”.
In recent days, prosecutors have moved to present evidence they say will prove Besigye and others plotted to overthrow the government
Museveni, 81, was declared winner of the last election in January although the results were rejected by runner-up Bobi Wine, who has since gone into exile in the United States. %!s()
How much of what you see on social media comes from sources that repeatedly post falsehoods? And if that content quietly vanished from your feed, would you believe anything different months later?
… %!s()
A new study published today in Science Advances offers the most rigorous answers yet. It comes from a collaboration between Meta and academic researchers who studied the 2020 U.S. election. It was run from inside the machine—on the live feeds of Facebook and Instagram.
The answers sound reassuring: Exposure to lies was rare for most users, and removing them did not measurably shift anyone's beliefs. But, as a researcher who studies how misinformation permeates these platforms, I think the fine print matters more than the headline.
Harder than it looks
Our political attitudes are shaped by television, talk radio, politicians, family and friends, so isolating social media's contribution is particularly difficult. People also self-select: Those drawn to fringe content surround themselves with others like them in tight clusters, so observation alone can easily miss them and prove little.
The clean way through is a randomized experiment inside the platform, and only the platform can run one. That is what makes this study valuable.
The team of researchers, led by Olivier Bergeron-Boutin from the University of California, Berkeley, first measured how much content—across 231 million adult U.S. Facebook accounts and 200 million Instagram accounts—came from "untrustworthy sources." These included pages, groups, accounts and websites repeatedly caught posting falsehoods by Meta's third-party fact-checkers.
Exposure was rare on average: 1.1% of what the median Facebook user saw and 0.1% of what the median Instagram user saw. But it was strikingly concentrated. For example, some 53 million Facebook accounts—23% of users—received nearly 80% of content from untrustworthy sources.
For the most exposed 2.5%, untrustworthy sources supplied the majority (roughly 60%) of the political and social content they saw from pages and groups.
Tweaking the algorithm
For three months around the 2020 election, posts from these repeat offenders were removed from the feeds of half of nearly 16,000 consenting users. Exposure in the targeted feeds fell by about 70%.
The researchers tracked 10 outcomes on each platform, including belief in false claims, the ability to tell true from false, polarization and trust in media. Nothing moved, not even among the heaviest consumers (though estimates for that group are less precise).
So is the misinformation debate settled? Not quite. The weak points of the study sit at the joints between what was measured and what will likely be claimed on its behalf.
A source counted as "untrustworthy" only after Meta's fact-checking partners caught it twice. Those partners published roughly 300 U.S. fact-checks in January 2020; matching software then spread such verdicts across far more content, about 180 million labeled pieces during 2020.
But anything the pipeline didn't catch counted as trustworthy. "Exposure was rare" really means "exposure to detected repeat offenders was rare."
The gray zone stays invisible by design.
In the one domain where it has been quantified—COVID vaccine content—misleading material that was never flagged had an estimated 46 times the total impact of everything the fact-checkers caught.
And in the case of state-backed influence operations that weaponize information, accounts are rotated and stay below exactly these detection thresholds.
Not normal Facebook
The control group was not experiencing everyday Facebook.
Around early November 2020, Meta deployed 63 emergency election measures aimed at reducing the flow of falsehoods for all users. One reanalysis of a companion study estimates these cut the share of untrustworthy content in ordinary feeds by around 24%. The experiment measured extra cleaning on top of an already scrubbed platform.
The dose was modest too: About three of 250 daily posts a Facebook user viewed were removed (12 of roughly 1,050 on Instagram) for three months. This is compared with attitudes developed over decades and anchored in identity and community.
And, crucially, participants had to volunteer for a study. People who agree to participate in such experiments are not always representative of the population at large.
Twelve of the 30 authors are current or former Meta employees, and every measurement instrument is in Meta's proprietary pipeline. The lead academics took no money from Meta and had final say over the text, but the project's own independent evaluator previously called the arrangement "independence by permission" and warned it should not become the model.
Remembering what was actually measured
Since 2020, much has changed on social media—especially when it comes to how Meta manages misinformation and disinformation.
The company ended third-party fact-checking in the United States in April 2025. By the study authors' own admission, the intervention they tested "is no longer possible" there, and exposure to falsehoods "could increase considerably."
There is a risk this study will be leveraged by big tech platforms to exonerate themselves—for example, by saying "users are rarely exposed to" falsehoods. It may also motivate them to stop moderation efforts on the basis that "removing misinformation changes nothing."
But remember what was actually measured: detected sources, a scrubbed baseline, a small dose, a short window, and people willing to be watched.
The study is good science. But some conclusions drawn from it might not be.
Publication details
Olivier Bergeron-Boutin et al, Untrustworthy sources on Facebook and Instagram in 2020: Concentrated exposure but no attitudinal effects, Science Advances (2026). DOI: 10.1126/sciadv.adz6502
Swati Mestri holds a bachelor's degree in Electronics Engineering and has worked as a content editor since 2019. She has experience editing research documents across technology, health care, and materials science, and has a particular interest in technology and space.
Full profile →
Citation:
Research finds exposure to 'untrustworthy' social media is rare, the fine print is less reassuring (2026, July 30)
retrieved 30 July 2026
from https://phys.org/news/2026-07-exposure-untrustworthy-social-media-rare.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.
%!s()
New technology is allowing Ukraine's drone operators to pilot drones hundreds or thousands of miles from the front lines, but for now at least, operations still need soldiers to set up the drones closer to the danger.
Maks Muravsky/Global Images Ukraine via Getty Images
…
New technologies are allowing Ukrainian drone operators — who are high-priority targets — to control drones from relatively safer locations hundreds or even thousands of miles away. But human soldiers often still have to go to risky locations to transport, prepare, and launch these aircraft.
Ukrainian interceptor drone and other uncrewed aerial system operators are increasingly using satellite connections that allow them to operate far from the launch site — potentially even from another country. The makers describe these capabilities as a key way to protect the pilots in Russia's crosshairs.
The aircraft, however, often still need to be brought to a launch site and prepared, meaning soldiers are still needed closer to the danger, at least for now.
Ukrainian drone maker Wild Hornets, for instance, has developed technology that allows pilots to fly the company's interceptor drones from hundreds of kilometers away. The company told Business Insider that ground forces still transport and prepare the aircraft, though.
With satellite communications technology for combat drone operations, "we are able to put our pilots in any place when there is a stable internet connection for them," Dmytro "Liber" Zhluktenko, a former drone operator who now serves as a lessons-learned analyst with Ukraine's 413th Unmanned Systems Regiment "RAID," told Business Insider.
"But this doesn't remove the need for the ground groups, people who actually attach the explosive, put the drone outside of the bunker, et cetera," he said.
Ukraine typically still needs soldeirs closer to the danger to do tasks like attaching explosives to the drones.
Kostya Liberov / Libkos via Getty Images
Remote piloting removes the operator from the launch point, but not necessarily the rest of the support crew. Troops may still need to transport the drones, unload and assemble them, attach explosives and communications equipment, and position them for launch. There are efforts underway to leverage other uncrewed systems, such as ground robots, for launching interceptors and first-person-view drones, but these robotic motherships aren't yet ubiquitous along the front lines.
New technology is moving some people to safer locations, but protecting everyone involved in a drone operation by keeping them far from the fighting is "impossible so far," Zhluktenko said.
Trying to get troops out of harm's way
Ukraine is doing what it can to keep as many soldiers as safe as possible, but there are limitations. Forces are still needed at the front for a range of missions even as robotic systems become increasingly prolific.
On more traditional missions with drones controlled by radio frequency connections, operators generally need to remain closer to their drones. But units are trying to, when possible, move "our drone operators further to give them more safety," Zhluktenko said.
They do this "even if it comes at the expense of our capabilities, because these are our people and we value them so much."
"The first principle, the first rule, is the safety of them, but it's a very tough balance. We want to keep them extremely safe, but, at the same time, there is some work to be done. So it's a very tough, tough balance thing."
Keeping drone operators safe is critical for Ukraine because Russia often treats them as top targets, according to soldiers, drone operators, and military officials.
Ukraine's drones destroy more Russian targets than any other weapon, and pilots are force multipliers. One operator can fly repeated missions to collect intelligence and launch attacks.
That makes technology that separates pilots from launch sites particularly valuable.
Then-defense minister Mykhailo Fedorov said in April that "the pilot is no longer tied to the position. The drone is in the sky, while control takes place from a protected environment in Kyiv, Lviv, or even abroad."
"This increases interception effectiveness, minimizes risks to operators, and makes it possible to scale capabilities without being tied to the front," he said.
Ukraine's drone operators are seen as Russia's No. 1 target, so Ukraine wants to keep them as far from danger as possible.
Oleksandr Magula/Suspilne Ukraine/JSC “UA:PBC”/Global Images Ukraine via Getty Images
Ukraine is investing heavily in technology designed to keep its soldiers farther from the intense fighting, including drones, ground robots, AI, and autonomous systems that enable weapons to detect, target, and strike with little to no human involvement.
Ukraine wants to replace certain human tasks, such as front-line logistics, with fully robotic systems, both to protect lives and to offset Russia's manpower advantage.
Ukraine's defense ministry has said, for instance, that "the primary task of ground robots is to bolster our units and replace soldiers in the most dangerous areas."
President Volodymyr Zelenskyy made that goal explicit in April when he said Ukraine's ground robot force had carried out more than 22,000 front-line missions over the previous three months.
That meant "lives were saved more than 22,000 times when a robot went into the most dangerous areas instead of a warrior," he said.
Then Commander-in-Chief Gen. Oleksandr Syrskyi said in October that Ukraine was prioritizing "high-tech systems that will help reduce the presence of soldiers directly on the battlefield."
Ukraine's ultimate goal is "to have less people on the direct battlefield and at direct risk, Zhluktenko said. For now, at least, human beings are still in the mix. It's not total robotic warfare.
Read next
Sinéad Baker
You're currently following this author!
Want to unfollow? Unsubscribe via the link in your email.
Sinéad Baker is a Military and Defense Correspondent based in Business Insider's London bureau, writing about Russia's invasion of Ukraine and NATO actions.Sinéad most often covers soldiers' experiences, military strategy, battlefield developments, the defense industry's response, and geopolitical decisions that surround the war. She has reported from NATO’s frontlines and around Europe, has interviewed multiple prime ministers and defense ministers, has appeared on BBC News and The Guardian's politics podcast, and has been cited by Congressional hearings.Sinéad has also extensively covered US politics and previously led Business Insider's breaking news coverage from London.Sinéad previously completed a master's degree in investigative journalism at City, University of London, and has written for The Guardian, The Observer, and TheJournal.ie. Sinéad is the former editor of the multi-award-winning The University Times in Dublin.Expertise
Experiences of soldiers in Ukraine, including battlefield developments and tactics
Western military responses to the war, and lessons they should learn
New weaponry built for and in response to the war
Popular articles
Russia's invasion is showing the West war isn't just about having the best weapons — it's also about having more
Ukraine's Patriot kills of Russian planes and missiles have turned a US air-defense weapon with a troubled past into a hero
Ukraine's electronic warfare fight against Russian drones is so chaotic that its own are getting caught in the crossfire
Ukraine's defense industry says the fight against Russia has shown it that the West's approach to weapons is wrong
American sniper in Ukraine says he couldn't care less about expensive weapons like tanks. 'Give me artillery' and ammo, he says.
It isn't neglect or laziness that erodes a sharp mind — it's the accumulating stock of things you've already gotten very good at.
getty
Most accounts of a mind going dull revolve around the same explanation: disuse. If you stop reading, stop learning, coast — and the machinery rusts. Which is why the standard prescription is effort. Do the crossword, learn the language, keep the thing busy.
… %!s()
But the more common route to a duller mind doesn’t run through neglect. It runs through competence. Every year of experience adds to your private stock of “moves” that already work — the argument that you know persuades, the fix that almost always clears the bug, the standard way to handle a difficult client. That stock is genuinely valuable. It’s also the thing that quietly makes thinking unnecessary.
You can all this tendency “running on proven moves.” This includes solving today’s problem with last year’s solution, and explaining an idea with the wording that landed last time. From the inside, this never registers as decline. It registers as getting good at things.
How This Habit Can Block Creative Problem-Solving
Psychologists have a name for the trap: the Einstellung effect. This is when the first solution that comes to mind blocks a better one from arriving.
The clearest demonstrations come from chess, where strong players shown a position containing both a famous checkmate pattern and a faster, less familiar win tend to find the famous one and stop looking for anything better. Eye-tracking shows they keep scanning the board while sincerely reporting they’re searching for alternatives — their gaze simply never leaves the squares belonging to the solution they already have.
That last part is the whole problem. Introspection doesn’t detect it. And a 2022 study published in Frontiers in Psychology suggests the narrowing isn’t something people arrive with, it builds over time. Researchers tracked where people looked while solving problems with a familiar method, and found that the split between those who’d eventually spot a better approach and those who wouldn’t wasn’t there at the start. It emerged the longer the familiar method kept working. Success is what narrows the search, and it narrows it gradually, without announcing itself.
How This Habit Can Block Understanding And Expertise
The same reflex has a verbal form, and it may be the more common of the two.
If you say something well enough, it becomes a set piece. So the next time the subject comes up, the reasoning doesn’t get rebuilt, the paragraph gets you saved gets retrieved. Repeat that for a few years and you become excellent at delivering the summary while the derivation underneath it slowly goes unmaintained.
Rozenblit and Keil called the resulting overconfidence the illusion of explanatory depth: people believe they understand things in far more mechanistic detail than they do, right up until they’re asked to explain step by step, at which point their confidence collapses.
A 2023 study published in Judgment and Decision Making found the collapse doesn’t stay contained. Across three preregistered experiments, people who attempted to explain how a zipper works came away rating their understanding of unrelated things — how snow forms, how an earthquake happens — lower too. Struggling to explain one thing seems to generalize into a broader dose of humility about how much you actually know.
Fluency feels like understanding. It’s a poor proxy for it, and it’s remarkably easy to manufacture. A 2023 study published in Applied Cognitive Psychology found that people who searched online for an explanation rated their own explanatory ability higher afterward than people who were handed the identical explanation to simply read — and merely previewing search-result snippets, without clicking through to read anything, produced almost the same inflated confidence. The polish of an answer, not the work of producing it, is what convinces us we understand.
Why Only Failure Can Dismantle This Habit
The learning literature is blunt about what builds capability: difficulty. A 2022 study published in Nature Reviews Psychology reviewing over a century of research on spacing and retrieval found that trying to pull an answer out of memory strengthens it more than rereading it ever does, and that this holds up across ages, subjects, and settings, even though the strategy runs against how learning is supposed to feel, and most learners never adopt it as a result.
Related work points the same way. A 2023 study published in Educational Psychology Review, reviewing decades of research on guessing before being taught the answer, found that this kind of errorful generation — getting it wrong first, then seeing the correct version — generally beats being told the answer up front, so long as the correct answer follows. Even sitting a test on material you haven’t studied yet appears to improve how well you learn it afterward.
The pattern is consistent enough to reframe the whole idea of mental exercise. Struggle isn’t the price of learning. It’s the whole mechanism behind it.
Which changes what the crossword advice was ever worth. A puzzle someone is already good at is mental activity, not mental growth. What seems to matter isn’t being busy, it’s being in a position to be wrong. Attempting something where the outcome is genuinely uncertain and the proven move doesn’t apply.
The Caveat Of This Habit
None of this makes proven moves the enemy. They’re what expertise is. A doctor who reasons from first principles at every appointment isn’t sharper than her colleagues; she’s slower and worse. Fluent access to what works is precisely what frees attention for what’s actually new. Rederiving everything is paralysis, not sharpness.
So the question isn’t whether to rely on stored solutions. It’s whether anything in a given week still requires not knowing. Whether there’s a room where you’re the least experienced person, a problem the usual repertoire doesn’t cover, an argument that has to be built rather than recited.
For a lot of capable people, the honest answer is no, and has been for years — not through any decision, just through the ordinary accumulation of competence.
A mind dulled by neglect at least feels dull. A mind running on proven moves feels sharp, quick and effortless, because fluency and thinking produce the same sensation from the inside — and only one of them is still doing work.
Wondering whether you’re still genuinely growing, or just getting smoother at what you already know? Find out where you land if you have this habit with science-inspired test: Growth Mindset Scale%!s()
Silvija Martincevic is the CEO of Deputy, a global platform for managing hourly workers.
getty
I have been in tech long enough to remember when “learning computers” meant something specific and physical: sitting down at a beige desktop, opening a program with a friendly face on the box and training your fingers to find keys without looking down.
…%!s()
Mavis Beacon Teaches Typing was not assigned homework. Nobody made you buy it, but millions of people did anyway because they could see what was coming. The internet was arriving, and they understood something simple: You either learn this, or you fall behind.
That instinct, to move early when technology shifts, has repeated itself in every major wave of workplace change. And it’s happening again with AI.
We Have Been Here Before
In the late 1990s, typing wasn’t a “nice to have” skill; it became a survival skill. Work was going digital, whether organizations were ready or not.
Healthcare saw this early as hospitals were digitizing. HIPAA was coming. The American Nurses Association recognized informatics as a nursing specialty in 1992, not because technology was handed to nurses but because nurses were already reaching for it.
Then came mobile: a different disruption, a different skill set and the same pattern. Workers figured it out before most companies had a mobile strategy. Truck drivers were mapping routes on smartphones before logistics firms had apps. Nurses were texting patient handoffs before hospitals had secure messaging platforms.
The workforce adapts, as it always has. What has changed is the speed of the shift and how far ahead workers are moving relative to the organizations around them.
The current narrative around AI skills is off. Most discussions focus on technical roles like data scientists and engineers. The assumption is that AI adoption is a white-collar phenomenon and that frontline workers will follow later, but that's not what the data shows.
Deputy’s 2026 "Big Shift Report" found that nearly "75% of US shift workers say AI helps them leave work on time more often," a clear sign that the technology is delivering real value on the frontline. Yet adoption remains limited. Only 25% of workers currently interact with AI tools at work, and most have little visibility into how those tools are introduced or used.In fact, 80% say their employer does not clearly communicate its approach to AI. The gap isn't worker adoption. It's organizational readiness.
Who Is Actually Driving Adoption
Polyemployment has climbed to its highest level in over a decade in the U.S. Gen-Z makes up more than half of these workers, often balancing roles across hospitality, retail and healthcare. They now represent over 40% of the U.S. shift workforce, so this isn’t a niche group; it’s the operational core of the economy.
And they are not waiting to be trained. For someone juggling several jobs, AI isn't an abstract productivity concept; it's how they manage real-time complexity across employers, platforms and calendars. They're using it to move through shifts faster and make work fit their lives rather than the other way around.
The pattern is familiar. Workers move first, and systems catch up later.
What Mavis Beacon Got Right
It's worth pausing on why Mavis Beacon worked. It wasn't just that it taught typing but also that it made the skill tangible, measurable and immediately useful. You could see your words-per-minute improve. You could practice for 15 minutes and feel the difference. The feedback loop was tight, and progress was visible.
AI adoption today lacks most of those properties. The tools are powerful, but they're abstract. The use cases are broad, but they're not always clearly connected to specific roles. “Learn AI” is often presented as a general directive rather than a practical skill tied to everyday work.
Workers are filling that gap themselves by testing tools, sharing tips and applying what works. Without structure, that learning is uneven and hard to scale. The opportunity isn't to slow workers down but to match the way they already learn.
What Organizations Should Do Differently
The organizations that I see making real progress with AI are not starting with training programs. They are starting with the work itself, introducing tools at the points where work actually gets stuck: writing shift notes, filling scheduling gaps and handling routine communication. Workers adopt what solves a real problem in front of them. Abstract training doesn't stick. Immediate usefulness does.
The fastest adoption I've seen hasn't come from formal programs at all. It comes from one worker showing another: a shift lead who figured out an AI scheduling tool and passed it to her team, or a nurse manager who dropped a shortcut in a group chat.
Gen-Z is often helping older colleagues adopt AI in day-to-day work. The organizations making the fastest progress aren't the ones with the most elaborate learning platforms; they're the ones that make it easy for employees to use approved tools to solve real problems.
That starts with clear guidance: which AI tools are supported, what they're designed to do and how they fit specific roles. Seventy percent of workers say they're eager to realize AI's benefits. For many organizations, the biggest barrier isn't employee willingness; it's the lack of straightforward direction.
The Real Leadership Test
Every major technology shift of the last 30 years has followed the same pattern: workers adapt quickly when the value is clear. They did it with the desktop, they did it with mobile and they are doing it again with AI.
By some estimates, nearly 60% of jobs will require significant reskilling in the coming decade. That transition is already underway, not in training programs but in day-to-day work.
The workers of the 1990s didn’t wait to be told the internet mattered. They figured it out on their own, on kitchen tables, after long shifts and without a training budget. Frontline workers are doing the same with AI right now. The question is whether their employers will catch up.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
