Antwerp's 'human zoo': Belgium confronts shame of colonial past

Cover image: FOCUS © FRANCE 24

Issued on: 07/08/2026 - 
05:34 min

In 1894, when the Belgian city of Antwerp was hosting a World's Fair, 144 Congolese people were forcibly brought to the city to be exhibited in the Congo Pavilion. In this "human zoo", seven Congolese men ultimately died of disease. Today, this shocking episode still haunts Belgium as one of its colonial history's darkest chapters.


But in May of this year, the city of Antwerp began to confront this painful past by heeding demands from its Congolese community by unveiling a monument to the victims. It's a way to confront the demons of the wealthy Belgian port city that greatly benefitted from the colonial period. FRANCE 24's Alix Le Bourdon reports.

BY:

Alix LE BOURDON

Dave KEATING

Amazon behind massive private gas plant for new data centers

Washington (United States) (AFP) – Amazon confirmed Friday it is financing a massive, private gas power plant in Texas that could become the single largest source of greenhouse gas emissions in the United States.



Issued on: 08/08/2026 -  FRANCE24

The Amazon logo is displayed at one of its booths at the annual Consumer Electronics Show (CES) in Las Vegas, Nevada, on January 6, 2026 © Caroline Brehman / AFP

It is the latest example of tech giants going off the grid to get their AI operations online faster.

The news was first reported by market intelligence company Cleanview, which reviewed satellite imagery to connect three data center construction permits filed this week by Amazon to the gas plant, which is being developed by Pacifico Energy and is known as GW Ranch.

Previously filed permits for the plant show it would have 35 turbines and generate 7.65 gigawatts -- larger than any gas plant currently operating in the United States.

A bigger facility, at 9.2 gigawatts, is planned in Ohio under a public-private partnership involving SoftBank. That plant would connect to the grid as well as power a data center.

The Pecos County, Texas, plant would be permitted to emit more than 30 million metric tonnes of greenhouse gases a year -- more than any other single source in the country, though facilities rarely emit at their permitted ceiling.

Amazon said that by generating its own power, it would avoid passing the cost on to ordinary Americans, whose electricity bills have spiked as utilities charge consumers for the new infrastructure feeding the AI boom.

"Amazon believes in paying the full costs of powering our operations," a spokesperson said.

"Our new planned data center campus in Pecos County does just that: it's powered by new on-site generation that won't raise electricity costs for Texas families and designed to transition to grid-connected service as interconnection timelines allow."

The company added it was "exploring opportunities" for solar energy and battery storage on site, and it would make use of "non-potable, brackish groundwater" which it said would avoid putting pressure on the watershed.

Amazon added it remained committed to reaching net-zero carbon emissions by 2040 under The Climate Pledge, and had enabled 40 carbon-free energy projects totaling nearly 10 gigawatts in Texas.

Building so-called "behind-the-meter" power was once a niche practice, exemplified by Elon Musk's xAI deploying mobile gas turbines to its data center in Memphis in 2024.

xAI still accounts for most of the two gigawatts of behind-the-meter power online today, according to research by Cleanview, which estimates that around 97 gigawatts of such projects are planned involving every hyperscaler -- including Meta, Microsoft and Oracle, with enormous climate implications.

The precise figure depends on how many are built, how efficient the turbines are and how often they run, but the total emissions could easily top 200 million tons of carbon dioxide emissions a year, roughly equivalent to the annual emissions of more than 45 million cars, or about the entire carbon output of Pakistan.

"There are more efficient ways to use gas," John Rogers, an energy analyst for the Union of Concerned Scientists, told AFP. "Virtually none of these plants are proposing to use those, and that matters for on-site pollution, both direct and climate."

While less polluting than coal, gas plants are still significant sources of planet-warming emissions, and their operations, particularly upstream, also leak methane, the main component of natural gas and itself a powerful greenhouse gas.

Beyond their carbon footprints, they emit fine particulate matter, linked to a range of heart and lung problems; nitrogen oxides, which aggravate respiratory disease; and hazardous air pollutants including mercury, formaldehyde and benzene.

© 2026 AFP

EU steps up AI regulation as new transparency rules come into effect

The European Union is stepping up its oversight of artificial intelligence this week, as the main provisions of its landmark AI Act become applicable, giving regulators stronger powers to inspect advanced models and sanction companies that break the rules.


Issued on: 03/08/2026 - RFI

Cover image: New provisions of the European Union’s AI Act came into effect on Sunday, 2 August 2026, strengthening transparency requirements and giving regulators wider powers to inspect advanced models and sanction companies that break the rules. 

AFP - KIRILL KUDRYAVTSEV
01:27

Adopted in 2024, the EU AI Act has established one of the world’s most comprehensive frameworks for regulating artificial intelligence (AI). It seeks to protect consumers and fundamental rights while providing companies with clearer rules for developing and deploying the rapidly evolving technology.

Some parts of the law were introduced earlier. A series of prohibited practices, including certain forms of AI manipulation and exploitation, have been banned since February 2025, while obligations covering providers of general-purpose AI models began to apply in August last year.

From Sunday, however, the European Commission gained the power to enforce those obligations directly, including by conducting investigations and issuing fines.

New transparency requirements have also taken effect, obliging providers and users of certain AI systems to make it clear when material has been artificially generated or manipulated.

Greater transparency

The rules require AI-generated content to carry markings that can be detected electronically. People or organisations publishing deepfakes – highly realistic manipulated images, audio or video – must also disclose that the material has been altered or artificially created.

AI-generated or manipulated text published to inform the public about matters of public interest must generally be labelled, although exemptions apply when the content has undergone human editorial review and a person or organisation accepts responsibility for it.

Special provisions apply to artistic, satirical and fictional works to ensure that labels do not unnecessarily interfere with their display or enjoyment.

Systems placed on the EU market before 2 August have until 2 December to meet some of the new technical marking and detection requirements.

The legislation also places strict limits on biometric identification and facial recognition. Some uses are permitted for law enforcement under tightly controlled circumstances, but practices such as indiscriminate collection of facial images to build recognition databases are prohibited.

AI systems cannot use deliberately deceptive or subliminal methods to manipulate people in ways likely to cause significant harm. They are also barred from exploiting vulnerabilities linked to age, disability or a person’s social or economic circumstances.

Providers of powerful general-purpose models must maintain technical documentation, publish information about the material used to train their systems and establish policies to comply with EU copyright law.

Those developing the most advanced models – which may pose what the legislation calls a systemic risk – face additional requirements, including safety assessments, cybersecurity protections, incident reporting and measures to mitigate potential harm.
Stronger enforcement powers

The Commission’s AI Office, established to oversee implementation of the law, can now assess general-purpose models and investigate whether their providers are meeting the requirements.

Companies may be ordered to supply documents and information or grant regulators access to their models for evaluation. In some cases, the AI Office can examine a model before it is released on the European market.

National regulators will oversee other AI systems, including many smaller or more specialised applications.

EU investigators can require companies to take corrective action when they identify a breach. The Commission can also restrict the availability of a model or ask its provider to withdraw it from the market.

The expanded powers mark an important step for Brussels as regulators seek greater access to increasingly capable systems developed by companies based outside Europe, including Anthropic’s Mythos models.

Penalties will vary according to the seriousness of the infringement.

Companies using prohibited AI practices may be fined as much as €35 million or 7 percent of their total worldwide annual turnover – whichever is higher.

Breaches of other requirements can carry penalties of up to €15 million or 3 percent of global annual turnover. Providers of general-purpose models can also be penalised for supplying inaccurate or misleading information, failing to cooperate with regulators or refusing access to their systems.

Smaller companies will face proportionate penalties based on their size and financial circumstances.

Further rules will follow – from December 2027, stricter requirements will apply to high-risk AI used in areas including healthcare, employment, education, migration and security.

Rules covering AI embedded in regulated products such as medical devices, machinery, toys and lifts will take effect in August 2028.

The EU also plans to ban systems designed to create non-consensual sexually explicit images, including so-called nudification tools.

Social media giant Meta ordered to pay $567 million in New Mexico child safety ruling


A New Mexico judge on Thursday ordered Meta to pay more than $500 million, finding the social media company created a public nuisance and harmed children. The ruling follows a March jury verdict awarding the state $375 million after finding Meta liable for exposing children to online predators and other harms.


Issued on: 07/08/2026 
By: FRANCE 24

US tech company Meta is facing multiple lawsuits over allegations of children being harmed by its social media platforms. © Sebastien Bozon, AFP

A New Mexico judge on Thursday ordered Meta to pay more than half a billion dollars over the state's claims that the social media giant created a "public nuisance" and harmed children.

Meta was previously ordered to pay $375 million in damages to New Mexico in March when a jury found the company liable for endangering children by making them vulnerable to predators on its platforms, among other harms.

Cover image: Meta CEO Mark Zuckerberg arrives for a landmark trial over whether social media platforms deliberately addict and harm children, Wednesday, Feb. 18, 2026, in Los Angeles. © Ryan Sun, AP
01:55

"We disagree with the ruling and will appeal," Meta told AFP in a statement Thursday.

"We work hard to keep people safe on our platforms and have been transparent about the challenges of identifying and removing bad actors and harmful content," the tech giant added.

New Mexico Attorney General Raul Torrez sued Meta – which makes Facebook, Instagram and WhatsApp – in 2023, alleging the company failed to protect children from online dangers.

During closing arguments, prosecutor Linda Singer told jurors that Meta's algorithms had directed adults toward content posted by teenage users while the company concealed internal findings about the risks to young people.

Parental concerns

In March, the jury found Meta violated the state's Unfair Practices Act by misleading consumers about the safety of its products for children.

"This case has always been about protecting children, standing up for families, and making sure that one of the world's largest technology companies cannot profit from practices that endanger young people without consequence," Torrez said in a statement Thursday.

"Today's decision is a victory for every parent who has worried about what social media is doing to their child and every child who deserves to grow up safer online."

Around three-quarters of the new $567 million fund should be used for mental health treatment, and it's to be paid over five years.

The remaining amount is allocated across programs for "awareness and prevention", "screening and assessment", "referral, linkage, and coordination", implementation and evaluation.

Following the ruling, Meta said the company remains "confident in our record of protecting teens online and will continue to defend ourselves against claims that misrepresent the facts".

Meta will also be required to provide the court with updates twice a year regarding its efforts to comply.

Additionally, the ruling orders Meta to make changes to its products, including making continued efforts to keep children in New Mexico who are under the age of 13 off of Facebook and Instagram.

Age verification for social media users is becoming increasingly popular by regulators around the world, though implementing such policies has proven complex and difficult.

In his ruling, New Mexico Judge Bryan Biedscheid also noted that teenagers are particularly susceptible to features that are designed to boost engagement such as autoplay, infinite scrolling, like counts, push notifications and algorithmic recommendations.

Some of these design features involve industry-wide problems and free speech issues, the ruling says.

However, the judge ordered Meta limit push notifications and usage to 90 hours a month, around three hours daily on average, for users under age 18, in addition to hiding "like counts".
Bellwether trials

Thousands of lawsuits have been filed in the US seeking to hold social media companies liable for alleged harms to children and mental health.

More than 30 US states are suing Meta over similar allegations, with a potential trial set to begin in August in Oakland, California.

In July, a US teenager dropped his lawsuit against Meta just days before the trial was set to begin; the teenager also reached confidential settlements with YouTube, TikTok and Snap.

That case was considered a bellwether that could have set the tone for resolving thousands of similar lawsuits across the country.

The first bellwether trial for social media harms on mental health concluded in March, when a Los Angeles jury ordered Meta and Google to pay a 20-year-old woman $6 million.

In May, Meta, Snap, TikTok and YouTube reached confidential settlements with a Kentucky school district.

(FRANCE 24 with AFP)

AI before AI: The legacy of Norbert Wiener’s cybernetics

 (3/3)


British mathematician Norbert Wiener, who founded the field of cybernetics in the 1940s, may deserve a place alongside artificial intelligence founders Alan Turing and John von Neumann. FRANCE 24 revisits this overlooked visionary who was also one of the first to warn against the automation of society.


Issued on: 05/08/2026 -
FRANCE24
By: Sébastian SEIBT

Norbert Wiener, the father of cybernetics, not only influenced research into artificial intelligence but was also one of the first to warn of the risks associated with automation. © FMM graphics studio

The story of the founding fathers of artificial intelligence almost always features British mathematician Alan Turing and his computing machine. John von Neumann is often presented as the mastermind behind programmable computers.
Summer serie: AI before AI

Names from the Dartmouth group – such as John McCarthy or Marvin Minsky – are also sometimes cited as the “inventors” of artificial intelligence.

However, another figure is often overlooked or relegated to the category of second-rate pioneers: Norbert Wiener, the founding father of cybernetics in the 1940s.

ChatGPT, Claude and other AI chatbots can be regarded as the offspring of his ideas.

But it's uncertain whether Wiener – with his particular moral, political and scientific convictions – would claim any credit for these modern large language models.
A young prodigy

Wiener was born on November 26, 1894, in Columbia, Missouri, into a family of Jewish immigrants from Eastern Europe.

His father Leo Wiener, a highly strict professor of Slavic languages of Lithuanian origin, would “behave somewhat like Pygmalion and attempt to mould his son in his own image", said Pierre Cassou-Noguès, a philosopher at Paris 8 University and author of a fictionalised account of the mathematician’s life.

"At least, that is how Wiener describes it in his autobiography,” he added.

Thus, Wiener became “a young scientific prodigy moulded by his father”, agreed Mathieu Triclot, a philosopher specialising in the history of technology at the Belfort-Montbéliard University of Technology.

He learned to read before the age of four, graduated from high school at 11, earned a bachelor’s degree in mathematics at 14 and completed his PhD in mathematics at Harvard at 18.

A few years later, the Massachusetts Institute of Technology (MIT) appointed him as one of its youngest professors.

It was in this role that, after World War II, Wiener developed his ideas which would “have an impact on many fields related to artificial intelligence such as robotics, control engineering and multi-agent systems”, said Philippe Mathieu, an artificial intelligence specialist at the University of Lille.

Weinter laid the foundations for the new field of cybernetics as early as 1943 in a seminal article and further developed his ideas in the book “Cybernetics or Control and Communication in the Animal and the Machine”, published in 1948.

Cybernetics studies “certain phenomena of control and information transmission in the same way in humans, in the animal kingdom and in the world of machines”, Cassou-Noguès said.

In other words, for Wiener and “cyberneticists”, it is possible to draw parallels between the way human and animal brains and machines process information.
Cybernetics, a ‘super-science’ attracting the biggest names in AI

In practical terms, these principles inspired Wiener to work on a new kind of anti-aircraft defence system during World War II.

He aimed to create a system capable of adapting in real time to the movements of missiles or aircraft and of learning from its mistakes. These were the very first tentative steps towards what we would now call “machine learning”.

“The central theme of cybernetics is understanding how an entity adapts to its environment and to the information available. Intelligence is seen as the result of interactions between an entity [living or not] and its environment,” Mathieu said.

“Logic, mathematics and electrical engineering, information theory, brain research and psychology – a whole bundle of scientific disciplines influenced the events that led to AI," writes Rudolf Seising, who specialises in the history of science and technology at the Deutsches Museum in Munich.

"In the first half of the 20th century, interdisciplinary considerations and, above all, transdisciplinary approaches to AI were mainly found under the umbrella of cybernetics, a ‘super science’.”

This “jack-of-all-trades” aspect of cybernetics may also explain why Wiener is less often seen as the “father of AI” than Turing or von Neumann, who had a more direct and immediate impact on the development of computer science.

READ MOREThe Dartmouth Workshop: The $7,500 investment that gave birth to AI (2/2)

But for Triclot, cybernetics’ “indirect influence” on AI is considerable.

A series of ten meetings known as the Macy Conferences on Cybernetics brought together scientists interested in Wiener’s ideas between 1946 and 1953.

Some of the biggest names in the history of AI, including von Neumann and Claude Shannon, took part and identified themselves with cybernetics at the time.

One of the founding members of the Macy Conferences was psychologist Joseph Carl Robnett Licklider, “one of the central figures in the history of American computing”, Triclot said.

He later became the head of the US Defense Advanced Research Projects Agency. Licklider "played a key role in the funding and development of the internet and also took an interest in human-computer interface technologies in the 1960s”, Triclot added.

At MIT, Wiener also supervised the work of Walter Pitts, a researcher and logician who was pivotal to the development of modern AI.

In the 1940s, Pitts laid the foundation for the first neural network model.
Against the Manhattan Project

This work, which is rooted in cybernetics, is essential to understanding today’s large language models.

While cybernetics itself may now seem to be a relic of the past, it paved the way, alongside neural networks, for the emergence of the branch of AI known as “connectionism”, which lies at the heart of the rise of ChatGPT and other 21st-century AI systems.

Wiener not only provided an “intellectual cradle” for AI, but “he was also one of the first to warn of its dangers”, Cassou-Noguès said.

Known for his strong-willed character and notorious outbursts of anger, Wiener understood why certain areas of scientific research were dangerous. Consequently, he refused to take part in the Manhattan Project, which led to the development of the first atomic bomb.

Following the bombings of Hiroshima and Nagasaki, Wiener even published a famous open letter in which he stated his intention to no longer publish research that could be misused for creating weapons of mass destruction.

Albert Einstein supported the move at the time.

But with cybernetics, "as early as 1948" Weiner also foresaw “the possibility of a fully or almost fully automated society – a development which, in his view, could just as easily lead humanity to a world free from the constraints of work or the social hell of mass unemployment,” Cassou-Noguès said.

Wiener concluded that he should continue his work in order to influence the direction research in the field was taking.

Often labelled as politically left-wing – US authorities suspected that he was sympathetic to the Soviet bloc during the Cold War – Wiener was thus one of the first to warn of the risk of social upheaval linked to the advent of machines.

In this sense, “cybernetics is a profoundly political endeavour, preoccupied with the fear of technological unemployment caused by machines that could replace workers”, Triclot said.

Cover image: The Debate © France24
02:02




For him, it is “striking to read this as early as 1947”, a time when computers were still very rare.

But for Wiener, this risk would only materialise “if we fall into the trap of treating human beings inhumanely”, said Triclot.

In other words, according to Wiener, robots would only be able to replace humans if humans were reduced to having the status of robots by denying everything that distinguishes them from machines.

Although the term “cybernetics” seems to have fallen out of use, the ideas of its founder are relevant today, at a time when companies are citing the triumph of AI to justify layoffs.

This article has been translated from the original in French.

LA REVUE GAUCHE - Left Comment: Search results for CYBERSYN

The era of the machine: Silicon Valley shows off humanoid robots

Cover image: Focus © France 24

Issued on: 05/08/2026 - 
05:07 min

For decades, humanoid robots belonged to the realm of science fiction. But today, they're one of Silicon Valley's hottest markets. According to consulting firm McKinsey, automation could transform up to 30 percent of all hours worked across the US economy by 2030. Dozens of companies are already developing robots that can interact with people and even help with everyday tasks. FRANCE 24's Pierrick Leurent and Valérie Defert report, with Wassim Cornet.

BY:

Pierrick LEURENT

Valérie DEFERT

Wassim CORNET

Winemakers across Europe face the challenges of climate change

EXPLAINER


Winegrowers in France are harvesting their grapes unusually early this season because of the effects of climate change. While the rising global temperatures offer a new chance for winemakers in northern and Scandinavian countries, experts say it’s too early to expect a drastic shift in the viticultural map.


Issued on: 07/08/2026 -
FRANCE24
By:
Sonya CIESNIK/Sophian Aubin

Winemaker Alain Gleyzes (R) speaks with other farmers during a wine grape harvest in La Palme, southern France on July 30, 2026. © Lionel Bonaventure, AFP

After weeks of scorching heat, winemakers in France are not heading into the harvest season the way they have historically.

“Thirty years ago, they started the harvest at the end of September and would sometimes end at the start of November,” said Alicia Dorey, a French lifestyle and wine journalist.

“Now it’s beginning in August, so it’s worrying.”

Wildfires in southwestern France have already destroyed more than 40,000 hectares in the Gironde department this year. And the increasingly volatile climate conditions are threatening the wine industry, one of France’s most valuable economic sectors. Winemakers are adapting by harvesting early.


Elsewhere in Europe, rising global temperatures are helping a burgeoning winemaking industry in countries farther north, including Scandinavian countries. “This is why we find wine in England now and in Bornholm, an island off the southern tip of Denmark, among other places,” said Dorey.

Despite wine production appearing in unexpected places, experts say it's too early to expect a revolution in the industry.

“We cannot talk about a drastic shift yet,” said Hugo Seilly, senior lecturer at the Burgundy School of Wine & Spirits Business. Some vineyards in Norway and Sweden have reached economic sustainability by consistently producing quality grapes every year, he said. Yet Scandinavian entrepreneurs have been “driven by the size of their domestic market”, meaning they haven’t yet moved into exporting their wine.

While the total surface area that Spain dedicates to vineyards is decreasing, it still leads the world with 930,000 hectares (2.3 million acres) followed by France with 783,000 hectares of of vineyards. China is in third place (733,000 ha) followed by Italy (728,000 ha) and Turkey (410,000 ha).

But a bigger geographic shift in wine production from south to north could come later. “There are a lot of questions about the sustainability of vines in Spain, as most of their vineyards are drastically drier," Seilly said. "They suffer more than other countries from climate change and its consequences.”

Access to water is also a big issue. “There are projections that most of Spain will have a desert-like climate by 2050, potentially wiping away large areas of production,” he said.
France’s wildfires: Famous Bordeaux wine region at risk

Cover image: title 1 © France 24
01:31

The sweet spot

Meanwhile, winegrowers in France are doing everything they can to save their crops. “In certain regions, they are going to be harvesting at night instead of during the day. This is to avoid grape oxidation (a chemical reaction that occurs when the compounds inside grapes meet oxygen) in the hot sun,” said Jeffrey T. Iverson, a wine journalist based in Paris.

French producers this year will be “nervous about the exceptionally high levels of sugar and low levels of acidity that the grapes will have as they come into the cellar", Iverson predicted.

“This is going to create different problems as they are fermenting these grapes and trying to create a wine with a similar profile to what they have traditionally made,” he said.

Iverson said other innovative methods include “pruning their vines in new ways to create foliage umbrellas that protect the grapes, finding forms of natural sunblock like talc powder to protect the leaves from being burned [and] using cover crops to retain humidity and protect the soil microbes”.

Some producers have resorted to pulling up their vines and replanting them using drought-resistant grape varieties better suited to warmer climates.

Irrigation is usually prohibited, “to maintain quality standards in terms of structure and concentration of aromas in the fruit”, said Seilly. Yet some authorities have “unfortunately given in to the sirens’ chant and awarded the right to irrigate the vineyards to help the vine survive droughts like the one we have been experiencing across the country this year”.

The profile of French-produced wines is evolving as temperatures rise.

Viticulture has historically been possible between the 30th and 50th northern parallels, Iverson said. “France, at 41°N (in the south) and 51°N (in the north), has always been in that sweet spot. And the best wines tend to come from the northern reaches of those latitudes,” he said.

French wines have traditionally been characterised by higher levels of acidity and lower levels of alcohol. “With climate change, French wines are losing that finesse. These hotter summers are making the grapes ripen quicker, so you get higher levels of sugar, sometimes even [before] the aromas have a chance to develop. The levels of acidity also plummet.”

Winemakers are faced with a difficult choice, Iverson said: “either pick the grapes earlier to be able to retain that acidity or accept wines that are unrecognisable to what they made in the past”.

A vineyard near Båstad, in southern Sweden, August 7, 2026. 

© FRANCE 24, via Romain Chichery

Champagne in the UK?

Some French winemakers who depend on acidity are now investing in English vineyards. Once considered too cold a climate for wine, England has now warmed enough to be suitable for grape varieties used in Champagne.

Entrepreneurs even further north who venture into the winemaking business face their own unique set of problems.

Romain Chichery, a French winemaker who moved to Sweden with his girlfriend Emma Berto to produce wine there, said his adopted country would never be an eldorado for wine. “Researchers estimate that there are 4,000 hectares of viable land for producing wine. We could plant vineyards across this land, but it would be risky. You end up spending more energy maintaining the young vineyards than producing wine,” he said.

Chichery said newly planted vineyards are no match for the freezing Swedish winters. “Planting thousands of hectares of vineyards in the centre of Sweden would mean either setting ourselves up for failure or spending millions of euros to maintain the vineyards so they yield a decent production.”

Climate change will challenge wine producers across Europe in many different ways. “What is certain is that flexibility and expertise are going to be rewarded fully – and nobody knows better about both than independent winemakers,” Seilly said.

Cape Verde: Illegal sand extractions for construction industry erodes coastline

Issued on: 06/08/2026 - FRANCE24

In Cape Verde, almost nothing remains of the postcard-perfect image of certain beaches on the island of Santiago. For decades, the illegal extraction of sea sand for the construction industry has been eroding the coastline. This illegal trade now threatens ecosystems, contributes to the salinization of farmland, and forces women to risk their health just to survive.

Video by: Clémence WALLER

Saudi Arabia, Pakistan and Turkey sign mutual defence agreement


Saudi Arabia, Pakistan and Turkey signed a mutual defence pact on Friday that would treat any act of aggression on one of the three nations as an attack on all of them, the allies said in a joint statement. However, it did not specify if the agreement would bind them to any particular military action in each other's defence.


Issued on: 07/08/2026
By: FRANCE 24

Turkish President Tayyip Erdogan, Saudi Crown Prince Mohammed bin Salman and Pakistan's Prime Minister Shehbaz Sharif pose after signing a joint defence agreement in Mecca, Saudi Arabia, on August 7, 2026. © Saudi Press Agency, via Reuters

Saudi Arabia, Pakistan and Turkey signed a joint defence agreement in Mecca on Friday, wedding Sunni Muslim US allies alarmed at a regional conflagration that has brought Iranian missile fire onto Gulf oil exporters.

Iran and its allies have been firing on Saudi Arabia and other Gulf states, and blockading their energy shipments, since the US and Israel attacked it on February 28 in a major escalation after years of regional ​tumult.

The "Mecca Joint Defence Agreement" ‌is intended to strengthen collective deterrence against any act of aggression and stipulates that an armed attack against any of the three would be regarded as ⁠an attack on all, they said in a joint statement.

However, the statement did not give specifics on the commitments each had accepted and it was not clear how far their agreement would bind them to any particular military action in each other's defence.

Turkish Vice President Cevdet Yilmaz ‌said the agreement was not aimed at any specific country. A Turkish official added that it was purely defensive and did not abrogate any existing agreements ⁠the countries had with other states.

The three nations share concerns about the increasingly aggressive military stances of both Israel and revolutionary Shiite Iran, as their longtime US ally struggles to contain regional upheaval that has caused a global energy crisis.
Powerful political symbolism

"Three of the Muslim world's most influential states are convening at a moment of heightened ​uncertainty, demonstrating a growing willingness among regional and middle powers to coordinate more closely on security matters," said Abdulaziz Sager, chairman of the Gulf Research Center, ‌a Saudi-based think tank.

"The trilateral framework could strengthen the three countries' collective diplomatic and defence weight while contributing to the emergence of a more regionally driven security architecture," Sager added.

Turkey has NATO's second-largest military. Saudi Arabia is home to Islam's holiest sites and one of the world's top oil exporters, while Pakistan is the only nuclear-armed Muslim country.

However, Rayed Krimly, Saudi deputy minister for public diplomacy, said in a social media post that ‌the accord does not represent an effort to establish a military axis or a sectarian bloc.

"Nor is it linked to any nuclear ambitions or arms race," he said.

Although Pakistan agreed a mutual defence pact with Saudi Arabia last year, it has not responded to Iran's attacks on ​the kingdom militarily. It has also played down suggestions its nuclear deterrent might be extended to allies, saying last year of its pact with Saudi Arabia that atomic weapons were "not on the radar".

Both Pakistan and Turkey have also voiced support for a maritime defence coalition Saudi Arabia has announced to protect regional shipping and energy routes.

Friday's pact was signed in Mecca by Saudi Crown ​Prince Mohammed bin Salman, Turkish President Tayyip Erdogan and Pakistani Prime Minister Shehbaz Sharif and follows nearly a year of negotiations, reported by Reuters in January.
Pact builds on longstanding military ties

The Middle ​East has been in flames for nearly three years since the October 7, 2023 Hamas attack on Israel.

Israel has waged war ​in Gaza, invaded southern Lebanon and seized territory in Syria as well as launching two air campaigns against Iran. Tehran has fired on Saudi Arabia, Kuwait, Bahrain, Qatar, the United Arab Emirates, Oman, Jordan and Israel.

While Turkey and Pakistan, on the fringes of the Middle East, have avoided ​significant direct attacks, both are anxious to calm conflicts that threaten their own security and economies.

For Saudi Arabia, three years of Middle East conflicts have imperilled its oil exports and ambitious development plans and drawn big questions over the reliability of its longstanding US security umbrella.

Cover image: yemen AFP - KHALED ZIAD
01:28

Burcu Ozcelik, senior research fellow for Middle East Security at the RUSI think tank, said the agreement was focused on counterbalancing Iranian hegemony in the region and building deterrence against Israel rather than moving towards a confrontation with Tehran.

"I think the objective is very much the opposite," Ozcelik said.

Pakistan has provided training and technical assistance to Saudi forces for decades, ⁠while Turkey and Pakistan have exchanged warships and training aircraft. Saudi Arabia is a major purchaser of drones from Turkey.

Pakistan has deployed about 8,000 troops, fighter jets, drones and an air-defence system to the kingdom and opened negotiations with Kuwait to ⁠provide security in exchange for energy cooperation ​and investment.

Israeli Prime Minister Benjamin Netanyahu's office declined to comment. A spokesperson for India's Foreign Ministry said the pact was a development that the country was closely following.

Iran did not immediately comment.

(FRANCE 24 with Reuters)

Turkey, Saudi, Pakistan: Attack on one nation deemed an attack on all


Issued on: 07/08/2026 - FRANCE24


Pakistan, Turkey and Saudi Arabia have signed a mutual defence agreement, under which an armed attack on any one of the three countries will be treated as an attack on all. The pact comes as Saudi Arabia seeks to diversify its security partnerships after repeated attacks on its critical infrastructure amid the war in Iran. France 24 Correspondent Jasper Mortimer from Turkey shares further insights.

US appeals court halts construction of Trump’s $400 million White House ballroom project


A federal appeals court on Friday ordered US President Donald Trump’s administration to stop construction of its $400 million ballroom on the site of the White House's demolished East Wing. The court added that a president cannot fundamentally reshape the White House without congressional approval.


Issued on: 07/08/2026 - 
By: FRANCE 24

Construction work continues on a helipad and the White House Ballroom on the South Lawn of the White House on August 5, 2026, in Washington. © Tierney L. Cross, AP

A federal appeals court on Friday ordered the Trump administration to halt construction of its $400 million ballroom at the White House without congressional approval.

US President Donald Trump doesn’t have the unilateral authority to construct an 8,400-square-metre ballroom where the White House’s East Wing stood before he ordered its demolition, according to the decision by a three-judge panel from the US Court of Appeals for the District of Columbia Circuit.

Trump said in a post on social media that he would be appealing what he called an “unjust decision” to the US Supreme Court.

In a 2-1 decision, the US Court of Appeals sided with historic preservationists who sued to stop construction. The court said it would stay its own ruling for two weeks to allow the administration to appeal to the US Supreme Court.

“Whether or not a massive ballroom should be constructed is for Congress to decide and is not a matter for Executive self-help,” the court wrote.

“This holding has nothing at all to do with whether the proposed ballroom is desirable, or not, as a matter of policy. This ruling does not even necessarily mean that the Defendants may not ultimately construct the ballroom.

“What it does mean”, the court continued, “is that the Defendants may not do so during the district court’s expeditious litigation without securing Congress’s authorisation, as the Constitution and laws require”.

The president reacted to the ruling with a lengthy post on social media in which he quoted from a dissenting opinion issued by Judge Neomi Rao, whom he nominated. He also described the plans for the facility to include bomb shelters, a hospital and medical facilities, “Top Secret Military Facilities” and other security features.

“The Military and Secret Service are viewing this horrendous, politically motivated, and unlawful ruling as a National Security threat to our Nation in that the entire Complex is being built for the protection of our Country and, additionally, all future Presidents,” Trump wrote.

The panel upheld an April 16 order from US District Judge Richard Leon for the administration to halt above-ground work on the ballroom. Leon, who was nominated to the bench by Republican President George W. Bush, allowed for construction to continue on below-ground work on a bunker and other “national security facilities” at the site.

Trump, a Republican, railed against Leon’s decision on social media, calling him a “Trump Hating” judge who “has gone out of his way to undermine National Security, and to make sure that this Great Gift to America gets delayed, or doesn’t get built”.

The appeals court case was assigned to Judges Rao, Patricia Millett and Bradley Garcia. Millett was nominated to the court by Democratic President Barack Obama. Trump nominated Rao. Garcia was nominated by Democratic President Joe Biden. Both Millett and Garcia backed the decision, while Rao opposed it.

The National Trust for Historic Preservation sued to challenge the project in December 2025, a week after the White House finished demolishing the East Wing to make way for a ballroom that Trump said would fit 999 people.

Government lawyers have argued that the project – the biggest structural change to the White House in more than 70 years – includes critical security features to guard against a range of threats, such as drones, ballistic missiles and biohazards.

“These upgrades, alterations and improvements are essential to protecting the President, his family and his staff, as well as the White House itself, and the entire project flows from them,” government attorneys wrote in a court filing.

Washington or 'Trumptown': The architectural expression of a cult of personality

Attorneys for the preservation group countered that the lack of a White House ballroom doesn’t constitute a national security emergency.

“That commonsense conclusion is not clearly erroneous, and there is no basis to disturb it,” they wrote.

In an initial ruling that the government also appealed, Leon barred above-ground work on the ballroom from proceeding without congressional approval. In response, the appeals court instructed the judge to reconsider the national security implications of stopping construction.

In his most recent decision, Leon ordered a stop only to the above-ground construction of the planned ballroom. The judge stressed that the White House is free to proceed with below-ground work, including the construction of any bunkers, military installations and medical facilities.

“Defendants argue that the entire ballroom construction project, from tip to tail, falls within the safety-and-security exception and therefore may proceed unabated,” the judge wrote. “That is neither a reasonable nor a correct reading of my Order!”

On April 2, Trump’s ballroom won final approval from the 12-member National Capital Planning Commission, which is charged with approving construction on federal property in the Washington region.

(FRANCE 24 with AP)

Exclusive: WFP to extend Palantir contract despite concerns raised in leaked audit

INVESTIGATION

The UN's World Food Programme is finalizing a five-year renewal of its contract with US data software firm Palantir despite “key privacy concerns” raised in a leaked 2025 internal audit. The audit, seen by PassBlue and FRANCE 24, reveals UN misgivings about allowing Palantir to process data for the world’s largest humanitarian supply chain.


Issued on: 07/08/2026 
FRANCE24 

By: Jessica LE MASURIER / Dulcie Leimbach

Palantir © FMM graphics studio

The World Food Programme (WFP) manages one of the biggest biometric databases in the world. Its SCOPE system contains information on millions of people, many of them beneficiaries of global humanitarian aid.

To help the WFP track humanitarian aid and manage its supply-chain data, Palantir developed the Digital Operations and Transformation System (DOTS).

Founded in 2003, Palantir – named after the all-seeing eye in J.R.R. Tolkien's Lord of the Rings – was partly seeded with money from the US intelligence community. It has since developed AI-powered data-integration software that can collate, process and analyze massive amounts of data on governments, organizations or individuals – software that is now deeply embedded in the US military and surveillance complex.

Within the UN itself, there is a split between those who resist Palantir’s growing involvement in its agencies – who want to ensure that any adoption of related AI-powered programs is responsable and ethical – and the move-fast-and-break-things crowd trumpeting “efficiency”.


A cyberattack on the Rome-based WFP in May exposed sensitive data on thousands of aid beneficiaries in Gaza, underscoring concerns about data protection. However, the WFP said the attack was limited to its Self-Registration Application for Palestine, where displaced people can sign up for aid, but which has "no direct connection" to Palantir software.

Some WFP personnel have expressed concern that data from the UN agency on beneficiaries could be used to train Palantir's data systems. Some are also apprehensive about partnering with a software company whose technologies have reportedly been used by US authorities in immigration enforcement and by military forces for lethal operations in the Middle East, notably in Iran and Gaza. To quote CEO Alex Karp, Palantir's product is "used on occasion to kill people".

Despite the misgivings, a WFP spokesperson said the agency is set to renew its contract with Palantir.

A new report by independent researchers Tina Indrani Mason and Rose Worden for the "Tech for Bad" initiative says the WFP is still “investing heavily” in DOTS and warns of a “contravention of UN principles”.

Entitled “Palantir and the World Food Programme: The egregious partnership at the heart of the world’s largest humanitarian supply chain,” the report describes the UN’s partnership with Palantir as “a clear conflict of interest and a grave abdication of ethical responsibility”.

A screen grab of the WFP's DOTS program, powered by Palantir. © FRANCE 24

'Inadequate risk management'

Florida-based Palantir specializes in creating software that can sift through huge quantities of data and detect patterns or connections that would take human analysts weeks or months to uncover. But it underscores that it is not a "data-mining" company.

The leaked internal audit by the WFP’s Office of the Inspector General, dated August 2025, focuses on information and communications technology, highlighting “inadequate risk management” and a lack of “proper guidelines and policy” on data privacy in the WFP’s relationship with Palantir. The problem is listed as “‘high priority”.

"Since the inception of the partnership, key privacy concerns raised by both internal and external stakeholders regarding the use of DOTS have not been adequately addressed," the audit states.

The audit was completed as the original pro bono contract between WFP and Palantir was quietly extended.

“We are finalizing a renewal of our existing five-year agreement. The general scope and terms of the contract remain the same [as in the 2019 contract],” a WFP spokesperson told FRANCE 24 and PassBlue. This renewal is set to be signed in August, according to WFP sources.

The leaked audit states that the WFP “lacks a unified and trusted system” for managing its data, adding that “reputational considerations surrounding the partnership [with Palantir] have led to institutional uncertainty”.

It also raises questions about the potential costs and complications of ending the partnership. “At present, there is no clearly defined exit strategy for DOTS, nor clarity on the potential costs that may arise should the partnership be terminated,” the audit stated.

It notes that the WFP’s technology division “has assessed opportunities to export and save data from DOTS in case of the partnership’s termination”, but that such a plan has not been presented to key WFP or Palantir parties and “substantial investments” into DOTS continue.

This was not the first time Palantir involvement had been red-flagged by WFP; an audit from 2020, conducted by the agency’s legal office, also raised concerns. The team within the legal office that carried out due diligence audits was disbanded in June 2025 and such assessments are now entrusted to the WFP’s internal audit group, which reports directly to the head of the agency.

The leaked audit provided further details about the original 2019 five-year contract with WFP.

An expert in humanitarian aid familiar with the contract – who requested anonymity for fear of reprisals – told PassBlue and FRANCE 24 that it contains a provision that could legally allow Palantir to “train its software” on UN data.

“In the User Licensing Agreement they had specific language that WFP wanted put in to prevent retention of the raw data by Palantir, but then they had a sneaky line, which is, ‘Palantir reserves the right to utilise data for performance improvement of the system.’”

That line “would potentially allow Palantir to use the data to train algorithms”, the source noted.

The language appears similar to Palantir’s other user-agreement contracts, including its partnership with Britain’s National Health Service, according to Privacy International, a UK-based NGO.

Privacy International examined the details of Palantir’s 2019 license agreement with the US Department of Defense, which included the following clause:


“10. Usage Data. Palantir may collect analytics, statistics, metrics, or other usage data related to Customer’s use of the Products (i) in order to provide the Products to Customer; (ii) for statistical use (provided that such data is not personally identifiable); or (iii) to monitor, analyze, maintain and improve the Products.”

Palantir's business model is different from consumer AI companies in that it sells “enterprise software” – tech built for organizations to help them manage data.

The company's website states that customer data is not used to train its models unless explicitly requested, nor do third-party-hosted services have access to client information for use in training.

"Palantir does not use, retain or sell customer data for its own purposes," a spokesman said. "We do not use customer data to train models."

But if a contract contains a User Licensing Agreement that grants Palantir the right to use customer data for "performance improvement", it could potentially use that information to train or improve its own AI systems.

“Software providers have always wanted usage data to improve performance,” which helps them identify popular features, peak usage times and application failures. “And so they routinely ask for that right in their contracts,” said Chris Wlach, a New York-based technology lawyer.

“But today, data can also help train algorithms and AI models. So users have to approach these clauses more cautiously. They need to ask, 'What exactly is my ‘usage data’ and how, exactly, is it being used?'”

Former Palantir employee Juan Sebastian Pinto explained to PassBlue and FRANCE 24 how the firm could also use client data indirectly.

“Although Palantir doesn't allegedly train its systems on client data, it can develop use-cases/applications with that data and sell it to another client, potentially in a dual-use (weaponized) capacity," he said in a Signal message.

This could be problematic for the UN.

“The data that WFP and other humanitarian agencies have access to about some of the world’s most vulnerable people provides an otherwise unavailable training data set for understanding how civilian populations physically move during conflict, seek food, water and shelter, and provides insights to their mortality and morbidity,” said Nathaniel Raymond, the executive director of Humanitarian Research Lab at the Yale School of Public Health.

In the wrong hands, such information could be deadly.

“Humanitarian agencies are de facto stewards of both the personally and demographically identifiable information about the prime targets of multiple global militaries and intelligence services,” Raymond said.

The "Tech for Bad" report goes even further: “When hunger is ‘being increasingly exploited as a weapon of war’ WFP must reassess whether embedding the software of a complicit US and allied-military contractor within their data ecosystem – pertaining to the supply of food – compromises them.”

READ MOREExclusive: GHF ‘complicit in war crimes’ in Gaza, says former aid contractor

WFP told FRANCE 24 and PassBlue that it “undertakes an extensive due diligence process to ensure any services or capabilities provided adhere to WFP's strict data governance, privacy and security needs”.

“WFP processes and controls all its own data and has stringent safeguards built into all its technology contracts,” it added.
‘Hot potato’

The WFP first announced its partnership with Palantir on February 5, 2019. But the agency had been working on a pilot project with the company since 2017, around the time David Beasley, a prominent US politician and Trump admirer, took over leadership of the WFP.

Palantir confirmed it has been partnering with the WFP since 2017.

Enrica Porcari, WFP’s chief information officer and director of technology at the time, said that although the Palantir system had access to records of humanitarian aid distributions, there was no personally identifiable data about beneficiaries.

According to UN insiders, Palantir aggressively pursued access to SCOPE but a small group of WFP staffers brokered a deal in which the biometric data would be fenced off.

Nevertheless, the Syrian Red Crescent withdrew from SCOPE in 2019, fearing its data would be shared with the US software giant.

Details about WFP’s partnership with Palantir are not widely shared. One senior WFP official told FRANCE 24 and PassBlue that he signed a nondisclosure agreement.

“There was a deliberate choice, an unspoken rule not to communicate on Palantir,” according to Pedro Garbellini, a former communications officer in the WFP’s technology division. “Nobody wanted to deal with it due to the reputational risks.”

Garbellini, who worked for the WFP from September 2021 to December 2025, said there was a dispute between the communications teams from the technology and partnership divisions in 2024 over who would have to deal with Palantir-related issues.

There was an unwritten rule: “Never speak about Palantir.”

The subject was a “hot potato”, Garbellini said.

‘Why are we still partnering with Palantir?’

Palantir posted a social media manifesto in April extolling the virtues of US power and issuing a call to arms for the West in an era of AI-driven geopolitical competition. Critics described it as a "disturbing" vision that "embraces AI state surveillance of citizens".

Palantir’s co-founder Peter Thiel, a libertarian turned Trump donor, has explicitly stated: “I no longer believe that freedom and democracy are compatible.”

CEO Karp has famously called Palantir “completely anti-woke” and called for those undervaluing the company to be sprayed with “fentanyl-laced urine”.

At a departmental meeting at WFP headquarters in Rome in April, one employee who read the manifesto asked a crucial question: “Why are we still partnering with Palantir?”

The question was met with a round of applause from UN staff.

It was also a question that the acting director of the agency had been asked over dinner at the end of January, according to UN insiders.

A UN human rights expert, Andrew Gilmour, handed acting WFP director Carl Skau a letter signed by UN staff members from multiple departments calling on WFP to end its ties to Palantir.

The letter read, in part, that the staffers were concerned that WFP “may be in contradiction of its core human rights due diligence obligations” due to its relationship with Palantir. According to a source, Skau acknowledged that the relationship was problematic but said it would be hard for the WFP to get out of it.

WFP systems, he explained, were now so intertwined with Palantir that it would be expensive and arduous to move supply-chain logistics to another platform. Besides, he said, Palantir had helped the WFP save money and improve its efficiency.

Skau did not respond to FRANCE 24's questions about the exchange. Palantir declined to be interviewed but said in an email it is "proud to partner with the World Food Programme".
Funding fears

The Palantir contract extension comes as the Trump administration pushes an "America First" candidate, Luke J. Lindberg, as the agency’s new chief. Lindberg is a senior official with the US Department of Agriculture (USDA), which signed a $300 million deal with Palantir in April.

UN Secretary General Antonio Guterres has reportedly "slow walked" Lindberg's approval process, according to a well-placed UN source, and it is unclear whether the new WFP head will be named before the end of his term. The WFP is usually led by an American.

As the United States threatens to withdraw more funding from the UN, some senior WFP staff believe ending the agency’s partnership with Palantir could prompt retaliation from the Trump administration, according to UN insiders who spoke to FRANCE 24 and PassBlue on condition of anonymity because they aren’t authorized to discuss the matter publicly.

A 34 percent reduction in WFP funding from major donors in 2025 – particularly from the US, the agency’s largest contributor – triggered major layoffs and restructuring, with plans to cut nearly a third of its global workforce this year.

While the $45 million contract the WFP signed with Palantir in 2019 is pro bono – enabling Palantir to get a tax-break for the full book value – it "is also a vendor lock-in contract", one WFP employee explained on condition of anonymity.

This type of agreement can make a customer dependent on a single supplier for products or services – making it expensive, technically difficult and legally onerous to breach.

The WFP told FRANCE 24 and PassBlue it is not locked into any single technology provider.

The staffer said a WFP assessment found that developing a similar system internally would cost an estimated $20 million.

But this could prove to be cheaper in the long run than the current Palantir partnership, the person said. For example, consultants for WFP who are trained to use Palantir’s systems charge much higher rates. And Palantir could start demanding payment for the use of its programs at any time.

WFP said that using the pro bono software nevertheless requires the agency to pay for third-party cloud infrastructure, hosting and additional support. The UN’s procurement site shows payments to Palantir totaling more than $10.5 million from 2017 through 2024.

Palantir UN procurement 2026 © Screen shot, ASR procurement by supplier

One employee with extensive knowledge of WFP technology described Palantir’s software as “mediocre”.

“It doesn’t bring anything new compared to what is already available on the open internet,” the person explained. “It’s like they put some makeup on a product that already existed.”

Another former WFP staff member who used Palantir software extensively said that around 98 percent of WFP operations use DOTS, which Palantir designed to connect data across the organization’s many fragmented systems.

They described it as “the best tool in the sector”, adding: “Nothing comes close to it. Everyone has access to it, you can build and ingest your own datasets, develop your own tools, and drive hyper-efficiency across the board. There is very little oversight if you’re a self-starter.”

They added that the platform significantly improves efficiency and can be used, for example, to optimize the entire supply chain.

Removing Palantir would “collapse” WFP’s supply chain, the staffer said.
Global reckoning

As Palantir’s worldwide influence grows, so does the global resistance to it.

An international grassroots movement, called Purge Palantir, has successfully lobbied several entities, including New York City’s public hospital system, to end partnerships with Palantir.

Britain’s National Health Service (NHS) has reportedly been put on notice to prepare for an eventual break from the company after coming under pressure from members of parliament. The NHS admitted in May that some Palantir staff could access identifiable patient data despite previous statements to the contrary.

Meanwhile, a deal between the UK's financial watchdog and Palantir has sparked fears that Britons' financial information could be shared with the Trump administration, given that US law allows the government to access US company data by issuing a FISA warrant.

France’s domestic spy agency, the DGSI, decided in June to end its partnership with Palantir and opt for a domestic competitor due to national security concerns, as has Germany's internal security service.

France's ambassador for digital affairs and AI, Clara Chappaz, told FRANCE 24 and PassBlue that “the DGSI's decision to stop partnering with Palantir was taken, in part, over concerns about data sovereignty and digital resilience”.

Given the mounting global pushback, many employees are wondering why the UN continues to partner with the controversial data software giant on delivering aid to the world's most vulnerable people.

This FRANCE 24 investigation was published in coordination with PassBlue. Damilola Banjo and Khatya Chhor contributed to this report.