This article discusses the challenges faced by traditional news agencies in an era dominated by social media and artificial intelligence (AI). It highlights how social networks initially promised democratized information but ultimately failed due to misinformation and manipulation. The piece now raises concerns about AI's reliability, noting that while these technologies can quickly summarize global events, they often include inaccuracies ("hallucinations"). The author argues for the continued importance of professionally produced journalism, exemplifying this with a detailed account of AFP’s on-the-ground reporting during a catastrophic glacial collapse in Nepal and Tibet that killed over 1,300 people.
Written locally by qwen2.5:14b on 2026-09-25,
using this article's own text rather than the other coverage of the
same event (that is the story summary below).
Story summary
Over the past two decades, news agencies have repeatedly faced claims about their obsolescence due to social media platforms like Twitter and Facebook. The initial optimism stemmed from events such as the 2009 Iran Twitter Revolution and the broader Arab Spring (2010-2012), where citizen journalists played a crucial role in disseminating real-time information. However, this enthusiasm was short-lived as it became evident that when autocratic regimes like those in Iran, Egypt, and Syria restrict internet access or imprison activists, the flow of independent reporting diminishes significantly.
The reality is that social media algorithms prioritize engagement over truth, often amplifying anger and outrage to boost user interaction. This shift has been exploited by malicious actors who manipulate these platforms to spread misinformation. Consequently, while social media initially seemed like a democratizing force in journalism, it eventually revealed its limitations and vulnerabilities.
Written for “Fake News And Manipulation” on 2026-10-05,
grounded in this article and the 0 other(s) covering the same event.
THE obituaries of news agencies have been written on many occasions over the past two decades.
asserted
obituaries → write → decades
First we were told we were slower than Twitter and that our global network of salaried journalists was redundant because every citizen was a journalist now.
Who needs to pay for ethically-produced fact-checked content when information from the source is flowing directly into our phones in real time?
We all remember the euphoria of the Twitter Revolution in Iran in 2009 and the Facebook follow-up Arab Spring (2010-2012).
asserted
We → tell → 2009
It was intoxicating to see the free flow of information on social networks from citizens on the front line.
asserted
It → see → line
THE obituaries of news agencies have been written on many occasions over the past two decades.
asserted
obituaries → write → decades
But we also all remember the brutal comedown.
asserted
we → remember → comedown
What happens when the autocrats shut down the internet or throw your citizens in jail?
asserted
autocrats → happen → jail
Who continues to document the story in Iran or Egypt or Syria for the outside world?
asserted
Who → continue → world
It turned out social media algorithms were not driven by a need to search for the truth.
asserted
algorithms → turn → truth
Silicon Valley worked out it would make more money promoting anger and outrage.
asserted
it → work → anger
It also turned out bad actors could manipulate these algorithms to promote lies.
uncertain
actors → turn → lies
Not surprisingly, untrained citizens were not equipped to help society navigate the pollution of our information ecosystem.
asserted
society → equip → ecosystem
Wind forward another decade and we are here again as we are presented with another ‘news agency killer’, the utopian future of Artificial Intelligence.
asserted
we → wind → Intelligence
ChatGPT, Claude, Gemini or Grok can scan the world’s journalism output in seconds and spit out five juicy paragraphs to tell you about anything that’s going on in the world.
asserted
that → scan → world
This is the so-called AI Overview — it is causing traffic to trusted journalism sites across the globe to crash and it rarely pays for the expensively-produced professional journalism it hoovers up.
asserted
it → call → journalism
And so again we are asked: why pay for your professional journalism when information is being delivered free in real time?
asserted
information → ask → time
Why waste money employing journalists to gather information that can be aggregated by our new chatbot friends?
asserted
that → waste → friends
Part of the answer lies in the little chatbot disclaimer which says the information in your AI overview may actually be wrong.
uncertain
information → lie → overview
Would you like your news with or without hallucinations?
asserted
you → like → hallucinations
Getting to the facts costs money, time, expertise and effort.
asserted
Getting → get → money
When a huge glacial collapse in late August triggered devastating flash floods along the Nepal-Tibet border leaving more than 1,300 people dead, AFP deployed a team of 11 journalists on the ground in remote parts of Nepal and India.
asserted
AFP → trigger → Nepal
Within an hour of police confirmation of the first major flooding three of our journalists based in Kathmandu were on the road.
asserted
three → base → road
They know the language, the terrain and the people.
asserted
They → know → language
They carried expensive cameras and drone equipment.
asserted
They → carry → cameras
One of our photographers made a 13-hour journey by motorbike and truck to reach the main disaster zone.
asserted
One → make → zone
He was then able to use his contacts to take an aid helicopter further into the disaster zone.
asserted
He → use → zone
His stunning drone images of destroyed villages were seen across the world.
asserted
images → destroy → world
It requires journalists to be on the ground, sometimes putting themselves in considerable personal danger, to bear witness to events.
asserted
It → require → events
It also requires journalists who know the country and can turn to a reliable source network.
asserted
who → require → network
While part of our Nepal team headed to the mountains, the rest remained in Kathmandu to sift fact from rumour as stories of dead and missing emerged.
asserted
stories → head → dead
Many of the first reliable words and pictures published by the global media came from agencies such as AFP and our colleagues at Reuters and AP.
asserted
Many → publish → Reuters
Our clients use the content without hesitation because they know it has been gathered in an ethically correct way according to the transparent charters and trust principles which guide the work of all our journalists.
uncertain
which → use → journalists
They know that we have employed professional journalists to gather information and that we invest resources in training and expertise to ensure the people we send into harm’s way are well prepared.
asserted
we → know → way
Clients pay for the product because they understand quality news-gathering is not cheap.
asserted
gathering → pay → product
They also understand we work to the same ethical guidelines from every dateline on the planet.
asserted
we → understand → planet
It makes no difference whether the content comes from Nepal or New York or Gaza, it bears the same stamp of quality.
asserted
it → make → quality
On the other hand, making stuff up costs nothing.
asserted
making → make → nothing
Side by side with real journalistic images from the Nepal disaster zone were reams of deep fake images showing artificially generated destroyed villages and bodies lined up for burial.
asserted
reams → show → burial
The tools that have given us AI Overviews have also given every human on the planet the capacity to create confusing and damaging false images that clog up our news feeds and often go viral on social media.
asserted
that → give → media
At AFP we employ over 100 journalists exclusively working to debunk and fact check this growing fake tide.
asserted
we → employ → tide
Pushing back on lies and manipulation has become an essential journalistic complement to searching for the truth.
asserted
Pushing → push → truth
…and 4 more, not listed.
The Aporia
