Technology

United States-based chipmaker Nvidia is investigating its chief executive officer, billionaire Jensen Huang, after he confirmed he had an intimate relationship with a woman accused of working for Chinese intelligence.

In an announcement on Sunday, Nvidia said the probe would be conducted by a “committee of independent directors” with outside counsel.

“While the board has no reason to believe that either national security or Nvidia secrets were compromised, the board’s inquiry will examine all available information,” Nvidia’s board of directors said.

It said the board continued to have “confidence in Jensen and his leadership team”. Huang will continue to lead Nvidia during the probe.

The investigation follows a report in The New York Times last month that Jensen may have been the target of a Chinese espionage plot.

China and the US have been in competition to dominate key tech sectors like artificial intelligence (AI).

Amid the AI boom, Nvidia has seen demand for its chips and programming interfaces explode, resulting in the company being valued in the trillions of dollars.

For years, the US has sought to restrict the export of advanced chips to China, fearing the Asian superpower could gain an edge over its technology and military sectors.

But on October 29, The New York Times said it had obtained exclusive information that a Chinese woman named Alice Huang, or Huang Le, sought to cultivate close ties with Jensen Huang in an effort to glean information about Nvidia’s technology.

The Chinese Embassy in Washington, DC, has denied any allegations of spying in the case.

The New York Times also reported that Alice Huang’s employer, Megaspeed, had become one of Nvidia’s largest buyers, purchasing some $4.8bn in equipment from the US firm.

Jensen Huang’s ties to Alice Huang, who is of no relation, spurred fierce bipartisan backlash in the US Congress in the aftermath of the report.

“For months, bipartisan members of Congress have raised concerns about NVIDIA’s advanced AI chips going to China,” Democratic Senator Elizabeth Warren wrote in a statement.

“NVIDIA’s Jensen Huang and NVIDIA’s board owe Congress answers. And we need to pass bipartisan legislation we have ready to stop advanced American technology from going to China.”

Republican lawmakers in the House of Representatives also called on Huang to appear and testify before Congress about the affair on November 18.

“It is vital that Congress determines the extent to which US national security has been placed at risk,” Republicans John Moolenaar, Brian Mast and Rick Crawford said in a statement.

In an email to Nvidia staff on Sunday, Huang had admitted to having an extramarital relationship.

“I was unfaithful to my wife. My actions hurt her and our children. There is no excuse ‌for what I did,” Huang said.

Nvidia has grown rapidly in recent years with the rise of AI, producing chips for large language models developed by companies like OpenAI, Anthropic, Meta and Google.

Huang has also been closely working with US President Donald Trump since the Republican leader returned to office in 2025.

In August of that year, the Trump administration announced it would take a 15 percent cut of certain Nvidia chip sales to China.

Trump reportedly bought as much as $1.6m in Nvidia stock and sold about $1m in August, ahead of November’s midterm elections.

After the midterm races last Tuesday, stocks for AI-related companies, including Nvidia, tumbled in reaction to widespread Republican losses.

Trump has pushed back against calls to regulate AI companies, but many Democrats during the midterms campaigned on placing restrictions on the companies, including on the construction of data centres.

Nvidia lost $450bn in market value in a single trading day last week, but it rebounded on Friday, regaining 4 percent of its losses.

Italy says civilian infrastructure in Europe is “under attack”.

A series of bomb threats this week grounded flights and put airports on high alert. No explosives were found, and the threats appear to be false.

But the disruption is not limited to aviation. Hackers have also targeted supermarket freezers in several countries, raising temperatures and spoiling food.

Together, the incidents are raising fears and questions about a wider campaign of hybrid attacks.

Who could be behind these threats? Are they connected? And can European nations come together to counter these acts of aggression?

Presenter: Dareen Abughaida

Guests:

Eric Schouten – Founder and CEO of Dyami Security Intelligence

Dennis-Kenji Kipker – Research director of the Cyber Intelligence Institute

Patrick Bury – Defence and security specialist at the University of Bath

Data centres face protests across Europe due to the massive amounts of computing power and water they use.

Protesters in the Netherlands have broken into a construction site for a new Microsoft data centre in Amsterdam, the latest action in a growing movement sweeping Europe.

Several hundred activists occupied the site in the Sloterdijk district on Friday, responding to a call for a two-day protest across Europe against data centres, which handle the large processing workloads associated with artificial intelligence (AI). Demonstrators clashed with police as some climbed a crane and set off flares.

Nathan Timmer, of the protest group Geef Tegengas – which translates as Push Back – told local broadcaster AT5 that demonstrators were occupying the site in an effort to stop construction.

“We don’t want it to be built. Nobody needs it,” Timmer said. “The only thing this will be used for is AI slop. AI warfare, data surveillance and committing genocide. We don’t want that.”

Pro-Palestinian activists also joined the demonstrations in the Dutch capital. By late afternoon police were escorting activists off the construction site, though some remained, local media reported.

Dutch police said they detained 150-200 people suspected of trespassing during the protest.

Microsoft said it was aware of the Amsterdam demonstration.

“We understand that people feel strongly about the project, and we respect their right to express those views peacefully,” a spokesperson said. “However, we do not condone any actions that are unlawful or that put people’s safety at risk.”

Data centres face resistance around the world due to the massive amounts of computing power and water they use. Activists argue that the boom in facilities serving cloud computing and AI is moving faster than public consultation and environmental oversight.

Dozens of civil society and grassroots groups have said they will join the “Stop Dirty Data Centres” events on Friday and Saturday in 15 European countries, including Ireland, Poland, Denmark, Finland, Spain, Italy, the Netherlands, France and the United Kingdom, organisers said.

In Germany, a dozen protesters gathered in front of a data centre on the outskirts of Berlin on Friday.

“We want to put a stop to this blind construction of data centres around Germany, regardless of any environmental concerns,” said protest spokeswoman Caro Weber. “This is really, really dangerous for our future.”

About 100 people also protested against data centres in front of the Polish parliament, representing communities around Warsaw where new centres are planned.

In Italy, the fast-growing data centre industry is facing an equally rapid rise in grassroots opposition, particularly in the northern Lombardy region, the country’s main hub for digital infrastructure.

A tech stock rout that began overnight in the United States has persisted in Asian trading.

Market authorities in South Korea paused trading on Friday as prices plunged across the region. Global markets have been rattled by a sharp fall in the trading price of tech stocks, which is driven by fears that the Democrats’ victory in the US midterm elections will increase regulatory scrutiny of AI and slow the country’s data centre build-out.

Japan’s Nikkei index slumped 6 percent, its deepest one-day drop since April 2025. Indices in South Korea and Taiwan plunged more than 8 percent after the opening bell, before cutting losses as authorities scrambled to stem the selling.

Wall Street’s volatility “fear gauge” soared overnight on Thursday and the tech-focused Nasdaq sank 5.5 percent, wiping $3 trillion off S&P 500 firms, as investors fretted that the Democrats could put the brakes on the AI boom being encouraged by President Donald Trump.

“It was a brutal morning for many investors here in Asia. We saw that the AI sell-off continues here across many countries,” Al Jazeera’s Katrina Yu reported from Beijing.

“Some of those major markets have fortunately recovered some of those losses, but all of them are still clearly in the red,” she said.

Analysts say the political jitters are hitting an AI trade already on shaky footing due to worries about returns on massive spending on related infrastructure.

“The selloff shows that AI stock prices had been built on rosy illusions about the future, and now that illusion is gone,” Dong Baozhen, chairman of asset manager Lingtong Shengtai, told the Reuters news agency.

“The US stock market has been the standard-bearer and backbone of the global AI bull run. Now that backbone has snapped. Can AI bull markets elsewhere hold up on their own?”

Authorities step in

Markets cut some of their losses by the time Asian markets closed, as authorities worked to stem the panic.

Trading on South Korea’s Kospi market was temporarily halted amid the sharp slide. Finance Minister Lee Hyoung-il also eased rules on retail investors’ leveraged investments.

In China, state entities stepped in to buy up equities in a bid to stabilise the market, said Al Jazeera’s Yu.

The Taiwan stock market closed down 4.5 percent, recovering somewhat from its early morning slump. For the week, the index clocked a 12.2 percent decline, its biggest since 1991, LSEG data showed.

The government of the island state, a key producer of chips, sought to talk up the prospects for the sector.

Demand ⁠for emerging technologies like AI, ⁠high-performance computing and ⁠cloud services is expected to keep expanding, driving ‌demand for tech companies, the economy ministry said in a statement issued on Friday.

Korean AI chipmakers Samsung Electronics and SK Hynix fell nearly 6 percent and 9.4 percent respectively, cutting steep losses from earlier. TSMC fell as much as 8.5 percent in Taiwan, before closing down 4.7 percent.

In Japan, Kioxia plunged 11 percent while AI investor SoftBank Group slid 11 percent.

New regulations

The selling indicates investors are starting to factor in increased opposition to data centres. Analysts say markets have yet to fully consider the cost such building work has imposed on communities.

Toby Walsh, the chief scientist for the AI Institute at Australia’s University of New South Wales, said he expected new regulations on data centres in many countries, including the US, Australia and the United Kingdom.

“The developers were building them in a way that wasn’t actually getting social licence,” Walsh told Al Jazeera.

“They were building them in places where they didn’t need to build. You don’t have to build data centres in urban areas. You can build them in the middle of nowhere.

“I think it’s actually for the good of the tech industry in the United States that this has happened,” he added. “We’re spending money a bit too quickly. And maybe this is the way to let that bubble deflate in a manageable way.”

Lawmakers from California are poised to lead two of Congress’ most consequential energy and environment committees, putting the Golden State at the forefront of efforts to scrutinize the Trump administration’s sweeping changes to federal climate policy.

Reps. Zoe Lofgren (D-San José) and Jared Huffman (D-San Rafael) have served as the top Democrats on the House Committee on Science, Space and Technology and the House Natural Resources Committee, respectively, since Republicans took control of the chamber in January 2023. With Democrats now set to control the next Congress, both are the likely choice for chair.

The committees oversee major areas of federal science and energy research, public lands, water, wildlife and natural resources policy. While neither directly controls federal agencies, their leaders hold significant authority over oversight and legislation.

Both committees have advanced bipartisan legislation under Republican leadership, such as the Fix Our Forests Act from Natural Resources chair Bruce Westerman (R-Ark.) and Calif. Rep. Scott Peters (D-San Diego). But they have also pursued efforts to reverse Biden-era environmental priorities — such as narrowing the Endangered Species Act — and advance President Trump’s domestic energy agenda, which favors oil and gas production.

Lofgren said the change represents a real opportunity to restore faith in science after two years of policies that have undermined it. The committee under Republican leadership launched a probe into the National Academies of Sciences over a report reaffirming the evidence for climate change.

“We’re not confused that climate change is going on. Neither are most Americans,” Lofgren said. “Ask any farmer in my district, and they’ll tell you climate change is going on.”

It’s also a chance to put California’s embattled environmental ambitions back in the congressional spotlight. The White House has challenged the state’s climate goals, slashing $1.2 billion in funding for its nascent hydrogen hub, moving to restart oil drilling off the coast and striking deals with offshore wind developers to abandon their projects and invest instead in fossil fuels.

“Much of what the administration has done the past two years has moved forward simply because there was no congressional check and balance,” Huffman said, pointing to a long list of executive orders, no-bid contracts and spontaneous announcements from the president. “I think you’re going to see a dramatic shift when we begin to have a check in the form of Congress that does its job.”

Though committees cannot prevent the president from issuing executive orders, they can hold hearings, subpoena documents, investigate whether agencies are complying with existing laws, and draft legislation to protect specific programs or funding.

This might not have prevented some of the administration’s most consequential actions — such as the payouts to offshore wind developers, which effectively killed three of the five projects off the coast of California — but would have brought greater public scrutiny and accountability to the process.

The same is true of the administration’s use of the Defense Production Act to restart oil pumping operations at the Sable Platform off the coast of Santa Barbara, which the White House said is a matter of national security.

“We are awash in oil and gas production,” Huffman said, adding that the president has “relied on fake emergency declarations that no one has questioned or challenged.”

Lofgren said she would look into the administration’s dismantling of scientific agencies and cancellation of research grants. The U.S. Department of Energy — whose research and development programs fall under the Science Committee’s purview — recently acknowledged in court documents that the termination of $7.6 billion in funding for clean energy projects was targeted at blue states.

On the legislative side, Huffman said he would pursue a Democratic permitting reform proposal aimed at addressing soaring energy costs. Lofgren said she would push to reauthorize the Energy Act of 2020 and renew federal programs established by the CHIPS and Science Act, which together boosted spending for emerging technologies, clean energy and supply chains.

She would also advance laws supporting fusion research, quantum computing, next-generation nuclear and geothermal technologies, along with greater guardrails around artificial intelligence. The Science Committee’s current chair, Republican Rep. Brian Babin of Texas, has also supported some of these emerging technologies, including fusion and quantum research.

In fact, Lofgren was optimistic that cross-party cooperation would prevail. Nearly all of the committee’s bills adopted this year passed unanimously, or close to it, although many never made it to the House floor, she said.

“I’m not going to compromise on oversight, and I know that the Republicans may not agree with us on some of the oversight issues,” she said. “But when it comes to legislation, we have a long history of bipartisan collaboration, which I anticipate would continue.”

In other election news

Energy affordability was top of mind for voters on Tuesday, with soaring electricity bills at the center of many high-profile national races. Residential electric rates in the U.S. are averaging 18.2 cents per kilowatt-hour, up from 15.9 cents when Trump resumed office in January 2025, according to the latest data from the U.S. Energy Information Administration. Many Californians pay much more than that.

This turned out to be a formidable weapon for Democrats, many of whom made energy costs central to their campaigns. An analysis from the nonprofit advocacy group Climate Power found that more than $475 million was spent on political ads about utility and energy bills in the first 10 months of this year. In the most competitive races, 32% of all recent Democratic ad spending went to ads mentioning energy affordability issues, compared with 12% from Republicans.

Data centers were perhaps the biggest bogeyman of all. In Michigan’s high-stakes Senate race, Democrat Abdul El-Sayed called for a pause on new data centers until stronger federal guidelines are adopted, and argued that they must pay to cover their own energy demand and not pass down costs to ratepayers.

His opponent, Republican Mike Rogers, was criticized for investments into tech companies that could benefit from the data center boom, and subsequently backed a one-year moratorium on new data centers in Michigan and said guardrails should be put in place to protect against utility price hikes.

The candidates did not agree on the reasons for high prices, however. El-Sayed touted 100% renewable energy as the best path forward, while Rogers said renewables are driving up Michigan’s energy and utility costs. El-Sayed won the state race Tuesday night.

In other climate news

Record hot ocean temperatures, higher-than-normal tides and powerful storms could converge amid a potentially historic El Niño. That’s a recipe for more coastal flooding, erosion and large waves in California this winter. My colleagues Rong-Gong Lin II and Sean Greene took a look at how this “worst-case scenario” for El Niño could play out.

But the news isn’t all bad. After a couple with a plot of land above Santa Clarita marketed it as a potential “glamping” site, they decided instead to sell it back to the local tribe that originally lived there. Read Noah Haggerty’s joyful story about how this 51-acre piece of land along the Pacific Crest Trail will go back to the Fernandeño Tataviam Band of Mission Indians.

And finally, the annual United National Climate Change Conference, COP31, will kick off next week in Turkey. It comes only weeks after U.N. officials reported that the world will not be able to limit global warming to 2 degrees Celsius (3.6 degrees Fahrenheit) above pre-industrial levels — a goal established during the 2015 conference in Paris. This year’s conference is expected to focus on implementing existing climate commitments, reducing emissions and accelerating the clean energy transition, among other topics.

This is the latest edition of Boiling Point, a newsletter about climate change and the environment in the American West. Sign up here to get it in your inbox.

On the Boiling Point Podcast this week, hear how Hollywood’s approach to climate themes is changing.

Months into Russia’s full-scale invasion of Ukraine, Palantir’s CEO Alex Karp became the first chief of a major Western Big Tech company to personally visit Kyiv.

President Volodymyr Zelenskyy said in June 2022 that he was “delighted” to receive the support of such a company in the fight against Russia “on the digital front line”.

The Florida-based software development firm’s partnership has paid off.

Ukraine has since become the first nation in modern history to systematically collect battlefield data, a crucial asset for training military artificial intelligence (AI) to recognise patterns and make predictions.

The trove of real-world combat data obtained from video and sensor feeds over four years of conflict is helping Ukraine monitor, predict and respond to Russian attacks with the help of Palantir’s products.

Yet observers say the data is also boosting the capabilities of the United States corporation whose technologies are implicated in human rights abuses, including in Israel’s genocide in Gaza, raising questions about the growing leverage that private tech corporations are acquiring over sovereign states.

A ‘strategic advantage’

Palantir was founded in the aftermath of the 2001 attack on New York’s World Trade Center to help the US defence industry build a single data environment from which to monitor and respond to incoming threats.

To reflect such a mission, the company was named after a stone in JRR Tolkien’s The Lord of the Rings that grants powers that border on omniscience.

Yasir Atalan, deputy director of the Futures Lab at the Center for Strategic and International Studies (CSIS), said Palantir makes “different forms of data talk to each other”.

In the military context, data from missiles, radars, the automatic identification system (AIS) and open sources is gathered on one screen, offering a “larger view of the battlefield”, Atalan said.

Similar solutions are available to governments for civil purposes, including border control, health services, tax and revenue – and mass surveillance.

With more data, systems use statistical learning to recognise patterns to make accurate predictions, enabling military systems to make split-second, automated tactical decisions.

Under this premise, large datasets based on real-life scenarios are coveted by companies aiming to outperform their competitors.

“Good battlefield data is extremely rare and hard to come by,” Elke Schwarz, a lecturer focusing on digital technologies at Queen Mary University of London, told Al Jazeera. Because of its scarcity, “those with access to [it] have a strategic advantage”.

Data originating from Ukraine may not be directly applicable to another conflict but is still crucial for training, Schwarz said.

“Building a system that correctly identifies a tank, for instance, requires a large volume of tank images, ideally in all kinds of weather conditions,” she explained.

Similarly, large volumes of behavioural data – information on how people react to a given scenario – are key in tracking, profiling and predicting human activity in mass surveillance systems.

Ukraine’s invaluable battlefield data

The pinnacle of the collaboration between Palantir and Ukraine’s Ministry of Defence came in January with the launch of a specialised software platform, the Brave1 Dataroom, which uses battlefield data to train artificial intelligence for military applications.

But prominent figures in Ukraine’s defence sector have questioned their government’s decision to share the country’s valuable asset with a private tech company under undisclosed terms.

“Ukraine has something that no other country in the world has,” Valerii Markus, a serviceman and former chief sergeant, wrote on Facebook. “We are paying with blood for data that someone else can convert into capital, international influence, and big money.”

The company maintains it does not acquire customer data.

In an emailed statement to Al Jazeera, it said that “customers own and control their data”.

Palantir’s software helps them “integrate and analyse their existing data sets,” the company said, adding: “Palantir never uses or retains customer data for its own purposes.”

Schwarz, at Queen Mary University London, said Palantir does not need to possess the data to benefit from it. “Rather, it works with the data that is made available,” she said.

Civil liberties and data protection

Palantir’s technologies, contracted by private clients and government agencies around the world, are known to have been implicated in human rights abuses.

Its core software platforms, ICM and FALCON, facilitated the administration of US President Donald Trump in detaining and deporting thousands of people, including asylum seekers, since January 2025. Amnesty International found the company aggregated massive amounts of data, built detailed individual profiles to track migrants and asylum seekers and assisted with deportations, contrary to its obligations to avoid causing or contributing to human rights abuses under international law.

The company has also assisted Israel in its war in Gaza, which a United Nations inquiry found bears the hallmarks of genocide.

Last year, UN Special Rapporteur Francesca Albanese found “reasonable grounds” to believe Palantir provided AI platforms used for battlefield data integration and automated decision-making in the war that has killed more than 74,500 Palestinians since October 2023.

In the United Kingdom, the National Health Service (NHS), police, military and local councils all signed multimillion-pound contracts with Palantir to integrate user data, but the agreements have been met with public backlash.

Nonprofit organisation Corporate Accountability launched a campaign against the 330-million-pound ($440m) contract between the NHS and Palantir to provide a new data platform. The watchdog warned the company is “at the heart of the dystopian marriage of Big Tech, the surveillance and detention industry, and rising fascism”.

Palantir in April published a 22-point summary of Karp’s coauthored book on X, which was widely criticised for promoting an authoritarian, antidemocratic worldview by questioning the efficacy of liberal democracy and arguing that some cultures produce vital advances while others are regressive.

Fatimah Shaikh, Corporate Accountability’s deputy campaigns director, said, “Palantir is not a corporation we can trust” to hold British citizens’ sensitive personal information.

“We should all be able to get medical care without fearing for our safety or privacy,” she told Al Jazeera.

Ilia Siatitsa, senior legal officer at Privacy International in the UK, said tech companies like Palantir blur the line between military and civil applications. They are also opaque about how they use customers’ data by virtue of their involvement in national security, she added.

Siatitsa argued the company could potentially use personal information to train its products – including those with a military application – without explicit consent from the users.

“If my data is going to be feeding into a war machine, I want to know about it,” she said. “I need to have the choice, not only [as a matter of] privacy, but also from an ethical point of view.”

In June, Britain announced it was reviewing the NHS’s data contract with Palantir after parliamentary pressure.

Growing dependencies

Private tech companies are supporting military and civil infrastructure, to the point where governments may become reliant on them to conduct military operations or run national services, experts say.

Anna Nadibaidze, researcher at the Center for War Studies at the University of Southern Denmark, said the balance of power has been shifting.

“We’re used to state armies, nations’ armies being the main players,” Nadibaidze said, adding that now those players are reliant on products supplied by the private sector. “So what if they choose to not supply this infrastructure any more?”

She argued Elon Musk’s SpaceX offered a valid example.

Ukraine has come to rely heavily on the company’s Starlink satellites for communications on the battlefield and effective warfare, but in late September 2022, during a pivotal push by Ukraine to retake Kherson from Russia, the tech billionaire suddenly provoked a communications blackout.

Amid tensions between Washington and Europe, several countries are now refraining from using Palantir’s products. France’s Prime Minister Sebastien Lecornu in June said domestic intelligence services would ditch Palantir’s products in an effort to avoid “strategic dependency”. Germany’s military also said it would no longer use the company’s products.

Spanish publication El Confidencial revealed in July that Spanish authorities quietly instructed state-backed companies to avoid signing new contracts with Palantir, citing concerns that sensitive national security information could be exposed.

Palantir’s reliability as a partner has also been called into question in Ukraine after the company finalised in September a strategic partnership with Nebius, a Dutch company founded as an offshoot of Russian technology giant Yandex.

Denys Shtilerman, chief designer and co-owner of Fire Point, a Ukrainian defence-tech company that develops and produces UAV and missile systems, said “the very same Palantir that works with Ukrainian combat data” had struck a deal with Yandex cofounder Arkady Volozh.

Volozh’s description of Russia’s war as “barbaric” has not eased Shtilerman’s concerns.

“Our combat data is not free raw material for Western technology corporations. It is a strategic asset of the state,” he wrote on X in a response to Kyiv’s collaboration with Palantir.

“Ukraine has paid for it with blood. We have no right to simply hand over the most valuable asset we possess.”

Supermarket refrigerators and freezers across Europe have been targeted by a cyberattack that raised the temperature of internet-linked units, destroying and damaging large quantities of food.

On Monday, fridges and freezers at more than 100 shops in several European countries ran hot. Cyberattacks on supermarkets began last month, when hundreds of French supermarkets were left with spoiled food and damage to units which had overheated, costing millions of euros.

So far, it remains unclear why hackers targeted the supermarkets, but attacks have spread across the continent and even to the United States.

Here’s what we know:

Which countries in Europe have been targeted?

On Monday, several supermarkets in Europe were affected, including the following:

• Dutch supermarket operator Delhaize said dozens of its shops in Belgium and Luxembourg were hit by an “IT incident”.
• French retailer Carrefour said about 15 of its 700 Belgian locations were hit by refrigeration problems.
• The Colruyt group, which runs Spar stores in Belgium as franchises, said five of its shops were affected.
• Swedish Coop said cooling systems at seven of its 800 stores were affected, with disruptions ranging from complete shutdowns to partial operation.
• Swedish furniture retailer IKEA, which also operates restaurants at its stores, said one site in Mons, Belgium, had been affected by a cyberattack.
• De Laet & Van Haver, the Belgian butchery chain, said the temperature in its fridges at its store in Hove, south of Antwerp, was raised as high as 18 degrees Celsius, ruining meat worth about 15,000 euros ($16,764).
• The food wholesaler Musgrave said a cyber-incident affected refrigeration systems in 13 of its SuperValu stores in Ireland.
• In October, hundreds of French supermarkets suffered cyberattacks, spoiling tonnes of food, causing fires and melting parts of refrigeration units.

What has happened in the US?

In August, freezers at a market in Fort Huachuca, an Army base in Arizona, went into defrost mode, spoiling all the frozen food they contained, according to a post on Facebook by the base.

The same month, the naval station in Newport, Rhode Island, reported on Facebook that its refrigeration system was affected, leaving only dry goods were available for purchase.

The Pentagon considered the freezer malfunctions on the military bases to be a suspected hack, the Reuters news agency reported late last month, citing four officials close to the matter.

One official said Pentagon investigators were communicating with the French authorities about the issue.

How are cyberattackers targeting fridges and freezers?

Hackers can access the controls of fridges and freezers which use internet software and raise their temperature in order to spoil frozen food.

In some cases, hackers have increased the temperature of appliances to as high as 50C (122F) in European supermarkets, causing them to melt or even catch fire.

Awais Rashid, a professor of cybersecurity at the University of Bristol in the UK, explained that many supermarket refrigeration systems which are remotely managed and, often by third parties, have not been given necessary software updates to prevent hackers.

“In line with established cyber security practices, these systems should be managed via secure networks. However, this is not the case in these incidents and they are accessible on the Internet,” Rashid told Al Jazeera.

“A remote managing party can access a web server on the controller to configure the system including temperature changes and maintenance. There are known vulnerabilities in several controllers (many of which are past their end of life),” he said.

Rashid added that vulnerabilities include “easily guessable” passwords, which would mean anyone could gain access.

Why are hackers doing this?

The motive behind the attacks is unclear, but authorities believe it may be a deliberate attempt to disrupt food supplies.

On Monday, the European Commissioner for Agriculture and Food, Christophe Hansen, said the attacks on fridges were a “deliberate attempt to disrupt” the bloc’s food supply and to undermine consumer confidence.

He added that while “food security is not currently at risk … the food waste and impact on businesses are real”.

Rashid said that while there are few details about who may be behind the attacks or what their motives are, it is likely that these are “the motive here is likely disruption”.

“This has financial impacts on the supermarkets but also societal impacts in terms of security of the food supply chain. Critically, this risks undermining trust in infrastructures on which people rely daily,” he said.

Who is behind the attacks?

On Monday, Belgium’s prosecutor’s office announced that it had opened a criminal investigation into the latest fridge cyberattacks. But so far, no one has taken responsibility.

Theories have emerged linking state actors to cyberattacks generally, with North Korea most recently accused of being behind last week’s $6.3bn Binance cryptocurrency heist, which forced the crypto group to freeze accounts.

European intelligence officials have also warned that Russia is escalating its shadow war on the continent amid the ongoing conflict with Ukraine, with cyberattacks among the methods it uses.

On Tuesday, the UK’s National Cyber Security Centre warned that Britain needs to be prepared for an artificial intelligence (AI)-powered cyberattack on its critical national infrastructure within the next two years.

“By 2028, it is highly likely that Russia, China and other nation states will routinely use AI-cyber to achieve their geopolitical goals, with the UK facing an increase in scale and speed of sophisticated cyberattacks,” the NCSC said in a report.

The report comes shortly after the UK launched its “Be Ready” campaign, calling on citizens to stockpile enough tinned food, bottled water and cash to last three days in the event of a cyberattack or national emergency.

Rashid said that as our infrastructure becomes increasingly digitised, the risk of disruption from cyberattacks, especially “cyber-physical infrastructures such as refrigeration systems”, has risen.

“However, we should also note that a lot of our digital infrastructures operate safely everyday because good security practices enable that. The world works because cyber security works effectively. So it is important that all businesses follow good cyber security practices,” he said.

Oakland, California – Protesters dressed as cardboard robots chanted and danced to a drumbeat as they protested against new data centres outside Oakland City Hall last month.

Their concerns largely focused on how the centres — which process large amounts of digital information — might strain the energy grid.

The result, they argued, would not only be environmental harm but also sky-high electricity bills for local households.

“Power prices have gone up so much in Oakland that some people are thinking of shutting off their power,” said Carlos Gutierrez, an activist for Communities for a Better Environment, an environmental justice group.

“And the city is planning to provide power to a data centre instead.”

Opposition to data centres is growing across the United States, as residents express concern about their rapid expansion.

More than 3,000 data centres already operate in the country, with more than 1,500 others in development, according to the Pew Research Center.

But the backlash is perhaps most striking in California’s San Francisco Bay Area, a region considered to be the heart of the artificial intelligence (AI) boom.

Major AI companies such as Anthropic, OpenAI and Meta are based in the Bay Area, and the industry’s growth has been credited with fuelling a local economic renaissance.

Data centres house the advanced computer chips and servers needed to run AI-powered services such as Anthropic’s Claude and OpenAI’s ChatGPT.

But they also require 24-hour power and a large supply of water to cool their processors.

Even with such cooling technology, data centres are notorious for emitting heat, leading to an increase in land surface temperatures as high as 9 degrees Celsius (16 degrees Fahrenheit).

A midterm election issue

Concerns about the cost of rapid data centre growth, both environmentally and financially, have catapulted the issue into the discourse around Tuesday’s midterm elections.

In the final two months before the vote, at least seven Bay Area cities, including San Francisco and Richmond, have passed temporary moratoriums on data centre construction.

They include Oakland, which passed a data centre moratorium on October 6, after days of protest.

Local politicians competing in Tuesday’s midterms have also used the topic as a wedge issue, to rally support among certain pockets of voters.

One of the most closely watched races in the Bay Area is for the seat left vacant by longtime Democratic leader Nancy Pelosi. After 39 years in the House of Representatives, Pelosi is retiring at the end of the year.

Two Democrats are competing in Tuesday’s election to fill her position in Congress: San Francisco Supervisor Connie Chan and state Senator Scott Wiener.

Each one has staked out a different position on AI data centres. Wiener, for instance, opposes any ban on data centres.

“A blanket data centre moratorium could undermine AI regulation,” he told the publication Mission Local.

Chan, meanwhile, has pledged to seek a nationwide moratorium on data centres, at least until responsible AI regulation is put in place.

“Let’s start with a moratorium on data centres. Allow us to have that moment to regulate artificial intelligence,” she said at a debate in October.

Chan played a leading role in passing San Francisco’s moratorium in early October.

A poll from the University of Southern California showed Wiener leading Chan 47 percent to 41, as of mid-October.

Navigating public backlash

“Right now, data centre opposition is an issue that unifies the US public like no other,” said Shum Preston, a political analyst based in San Francisco.

So far, there has been relatively little opposition to the data centre moratoriums in local hearings. But he speculates that there may be several reasons for the lack of pushback to the bans.

Among them are the surveys indicating widespread public backlash to data centre creation.

An October CNBC poll found that 63 percent of registered US voters oppose building data centres within 16 kilometres, or 10 miles, of their homes.

“Data centres poll very poorly, and developers probably don’t want to put their supervisor friends in trouble ahead of elections. So they probably let this pass,” Preston said.

He also signalled that the short-term nature of the moratoriums may be a factor.

Certain Bay Area cities like San Francisco, Brisbane and Oakland paused construction on data centres for 45 days, with the option to extend the ban later.

Opponents of the moratoriums “will be more vocal when” the 45 days end, Preston said. By that time, the heat of the midterm election season will also have settled, he added.

“That’s when the hearings will get more contentious,” he said.

Not long enough?

Another hurdle facing the moratoriums is the lack of transparency surrounding data centre construction.

Supervisor Shamann Walton, who sponsored San Francisco’s moratorium ordinance, has argued that the city’s definition of “data centres” is outdated.

Created in 2001 — long before the AI boom — the city planning code defines data centres as “internet service exchanges”, a definition that may not accurately characterise the technology today.

Walton said the outdated language has made it hard for authorities to count how many data centres exist in the city.

Not being able to identify data centres can also prevent city officials from monitoring their power and water usage.

“We cannot use 2001 rules to regulate 2026 technology,” he said in a September 14 statement.

Another issue is the time it takes to carry out the necessary environmental assessments. The moratoriums may not offer enough time for a comprehensive review.

“Forty-five days is hardly enough time to document and study the impacts and then frame regulations,” said Roger Lin, senior attorney at the Center for Biological Diversity, a US environmental nonprofit.

He added that legal barriers limit access to such information from the data centres themselves.

“Data centres often have nondisclosure agreements, so even local air quality boards may not know the impact of data centres on air quality,” said Lin.

Supporters tout economic benefits

Still, AI industry supporters have touted the growth of the data centres as evidence of a flourishing industry, with positive downstream effects for US communities.

Anthropic, for instance, is preparing an initial public offering (IPO) that is likely to raise billions of dollars.

But critics warn that the company’s growth may not translate into an economic boom for the San Francisco Bay Area overall.

Instead, residents fear unequal economic growth that primarily benefits the wealthy. Already, San Francisco faces some of the highest and fastest-growing home rental prices in the country.

At a hearing at San Francisco City Hall, a resident vented their frustrations about the data centres to the supervisors.

“This city is less for humans and more for robots and billionaires,” the resident said.

Russian Emergencies Ministry also says its mobile app suffered a ‘massive cyberattack’, leading to ‘false information’.

Ukraine’s Air Force has said its air raid alert app sent erroneous messages after allegedly being hacked by Russia.

Fake alerts warning of nuclear and chemical threats to areas in and around Kyiv were sent on Sunday, along with warnings of potential street combat. An erroneous message also said Russia’s hypersonic Oreshnik missile had been deployed.

“The air alert app appears to have been subjected to enemy influence,” the Ukrainian Air Force said on Telegram on Sunday. “There is no radiation threat! Stay calm and follow our updates on our channel.”

SPRAVDI, Ukraine’s Centre for Strategic Communications, blamed Russia for the hack in a post on X.

The raid alert app is used across Ukraine to warn people in real time of incoming Russian air attacks. A separate app, specifically for the Ukrainian capital, was working normally, according to Kyiv authorities.

Hours later, early on Monday morning, Russia’s Emergencies Ministry said its mobile app had been hit by a “massive cyberattack”.

The ministry said on Telegram that the attack led to the spread of “false information”.

Russia’s and Ukraine’s announcements follow the end of a truce between the two countries in Kherson.

The localised ceasefire, which began on Tuesday and ended on Sunday, was the longest of the war so far.

Earlier this month, authorities worldwide urged website administrators to execute emergency updates after a nonprofit programming foundation flagged critical security flaws that could allow hackers to breach systems.

Government and corporate websites briefly went offline to patch the vulnerabilities, nicknamed PHP2Shell, which were discovered by independent cybersecurity experts using artificial intelligence (AI).

The vulnerabilities were embedded in the code of a widely used open-source web programming language, PHP, for more than a decade. They had gone unnoticed, however, because it would have been too complicated for humans to exploit them without the use of advanced AI.

But cybersecurity experts, as well as those maintaining important software, say PHP2Shell could be the first in a deluge of security flaws exposed by emerging AI. They say much of the world’s most critical digital infrastructure is maintained by volunteers, who often lack the resources and time to respond to newly discovered threats.

“There’s been a lot of people in the AI safety community that said this is one of the first areas where we are going to see the harmful impact of AI,” said Peter Slattery, a research scientist at MIT’s FutureTech, who studies how cybersecurity experts are dealing with the new technology and tracks real-world incidents involving AI.

A digital world built on volunteer work

PHP got its start in 1995, when Danish-Canadian programmer Rasmus Lerdorf sought to track how many people visited his online resume. Over time, others around the world improved Lerdorf’s software, and it grew in popularity.

As with other open-source coding projects, anyone could download, modify and share improvements. Websites like Facebook and Wikipedia were built using PHP, which runs about 70 percent of the world’s websites.

In 2021, a dedicated nonprofit, the PHP Foundation, was launched to coordinate development. Like many other foundations maintaining open-source software, it operates on a meagre budget.

Most of the digital infrastructure consumers use is powered by open source software. The Mozilla Firefox browser is open source, as are languages like Perl, Python, and JavaScript. The software that lets a computer communicate with others on the internet is often open source, with many servers using products like Apache HTTP Server, maintained by the Apache Software Foundation.

“The world today, including proprietary software vendors, relies on open source software for its digital infrastructure,” said Amanda Brock, CEO of OpenUK, a nonprofit that supports open source software.

The world’s largest corporations, like Meta, Google, OpenAI and Anthropic, depend on that volunteer-built, open-source infrastructure. Now, those companies’ own AI tools are uncovering security flaws in the infrastructure at a pace that needs far more resources for groups like the PHP Foundation to keep up.

“It’s simply not possible with the resources in place to fix all of the known vulnerabilities,” Brock said. “Foundations are attempting to manage these challenges, but the smaller ones don’t have hundreds of millions in funding, and managing this at the pace AI requires is proving impossible.”

A deluge of new bugs to fix

The PHP2Shell vulnerabilities were found through several advanced AI models, which can employ far more complicated attempts to hack past security barriers than humans could.

Developers of other software are also discovering far more security flaws than in the past because AI is being used to probe defences.

Google said it uncovered more than 1,000 security bugs in the last update for its Chrome browser, more than in the last two dozen updates combined. Mozilla, the foundation behind the Firefox browser, found more than 400 security vulnerabilities in April, a significant jump from the previous record of 76 bugs discovered just one month earlier.

“We have maybe 10 times more issues than usual because we are using AI to find them,” said Jarek Potiuk, part of the volunteer security team at the Apache Software Foundation. (In keeping with the foundation’s non-hierarchical ethos, Potiuk said he has no real title).

Potiuk said there is an “industry-wide movement” to conduct scans using AI models and identify security vulnerabilities, much like was done to find the bugs in PHP. But the scans themselves cost money – used to pay for the tokens for AI models – and afterwards, humans must pore over them to weed out false positives.

“Every project, every foundation is doing this right now – spending enormous amounts of time to find these issues,” Potiuk said. “AI helps to find them, but you have to verify things, you have to review them.”

Earlier this year, Anthropic launched Project Glasswing, offering $100m in credits to support its advanced Claude Mythos AI model to search for security vulnerabilities in the world’s critical software. In one month, more than 23,000 vulnerabilities were found across more than 1,000 open source projects.

The Apache Software Foundation has taken advantage of another scheme, the Alpha-Omega project, to scan for security bugs across the 290 software packages it maintains. The project, backed by companies like Anthropic, Amazon, Google, Microsoft and OpenAI, spends about $7m a year helping open-source developers find critical security bugs.

But those working on finding and fixing the security flaws say the corporate world could be doing a lot more, given the trillion-dollar valuation of companies whose products would not work without free software like the packages the Apache Foundation provides.

“They are helping a lot, but we need more active players in this area,” Potiuk said, “because it’s highly disproportional – the impact it has on the corporate world versus the open source world”.

When a major security flaw like PHP2Shell is uncovered, it often comes down to a handful of volunteers to fix it.

Keeping the internet’s infrastructure as decentralised and open-source as possible is crucial, Potiuk said. Instead of buying out something like PHP, major players like Google or Microsoft could, for instance, pay staff to dedicate some of their work hours to maintaining open-source projects they use.

Real-world targets

The 2021 Log4Shell vulnerability, which allowed hackers to remotely run code on millions of devices through a widely used library maintained by the Apache Software Foundation, underscored the critical role open-source software plays in system security. That bug was discovered without the use of AI.

Suddenly, cars, microwaves and a host of electronic devices connected to the internet were vulnerable, Potiuk said.

“That was the moment when the whole ecosystem realised how important it is to take care of open source,” he said.

A bug like that, if found by a malicious hacker and exploited before there is a chance to patch devices, could have caused a major disruption.

Slattery said the database of incidents he and his colleagues maintain shows that an increasing number of vulnerabilities are being detected, and it is a matter of time before some are exploited in the real world.

“We are starting to see both a lot more vulnerabilities detected and addressed in time, but also attacks on a wide range of systems,” Slattery said. “We have been moving in this direction for a long time. There are more and more things now you can attack.”

Within a week, two of the three largest cloud computing companies in the world signed 20-year contracts for nuclear electricity. Amazon has taken 690 megawatts from Constellation’s Calvert Cliffs plant in Maryland, including 190 megawatts of additional generating capacity expected between 2030 and 2032. Google has signed an agreement that will bring 890 megawatts of additional nuclear capacity into the United States. Constellation is investing more than $4.3bn to upgrade 11 reactors across Illinois, Pennsylvania, and New Jersey. Neither agreement, however, involves building a new nuclear reactor. Rather, they are paying to generate more output from reactors that already exist – some of which are even older than the companies themselves.

A few years ago, such agreements would have seemed inconceivable. Nuclear power had fallen out of favour following accidents at Three Mile Island in 1979, Chornobyl in 1986 and Fukushima in 2011. Germany phased out nuclear power, while Japan shut down its reactors after Fukushima, although it has since begun restarting them. In many democratic countries, advocating for nuclear energy became politically difficult.

Today, an undamaged reactor at Three Mile Island is being brought back into service, with Microsoft contracted to buy its output for 20 years. The US Department of Energy is providing up to $4.4bn in loans to support the restart of three shuttered plants, with the aim of restoring about 2.25 gigawatts. Palisades in Michigan, which was closed in 2022, is contracted to be generating again by March 2027. If successful, it would become the first US commercial nuclear plant to restart after entering decommissioning.

What has changed?

The quiet reversal

In the last decade, the industry’s answer to AI’s energy problem was efficiency – things such as better chips, smaller models and smarter cooling seemed appealing and feasible. Industry conferences featured discussions about performance per watt, with modern artificial intelligence (AI) chips performing more computation per unit of electricity than their predecessors.

Yet those efficiency gains did not reduce overall electricity demand. Instead, each gain was quickly put towards something larger. Even when DeepSeek demonstrated that a competitive model could be trained at a fraction of the computing cost many had expected, the result was not a reduction in industry spending. Instead, laboratories continued planning more training runs, models that reason for longer, agents that work continuously in the background, and video generation.

This would not be the first time something like this has happened in history. In 1865, William Jevons observed that more efficient steam engines made Britain burn more coal, not less, because cheaper power created uses that had not existed before. The International Energy Agency estimates that global data centre electricity demand will more than double to around 945 terawatt-hours by 2030, even assuming continued improvements in efficiency.

Hence, the industry has changed strategy – unable to shrink demand, companies are now buying up supply.

The arithmetic of waiting

Money, however, cannot solve every problem quickly.

A new large reactor takes a decade or more from decision to electricity. Tech companies are investing in small modular reactors to power their data centres, although many of these projects are not expected to deliver electricity until the early 2030s. So far, very few of these have been built anywhere in the world. Thus, while reactor restarts and upgrades are the faster option, the “fast option” still means 2027, 2028, 2029, or beyond.

AI capital expenditure, however, does not operate on that same calendar. The chips being installed this year could be overtaken by newer technology long before the 20-year electricity contracts powering them expire. This discrepancy underpins all of these announcements, yet the industry has not addressed the implications if demand weakens while contractual obligations remain.

Who pays in the meantime

This is where the story stops being about technology.

Most of these deals sit inside PJM, the grid operator serving roughly 67 million people across 13 American states and Washington, DC. PJM uses capacity auctions to secure sufficient electricity-generating resources for future demand, with the associated costs ultimately passed on to consumers through their electricity bills. Capacity prices have reached the regulatory price cap in three consecutive auctions. According to the grid operator’s independent market monitor, data centre demand contributed approximately $29.4bn in capacity costs across the four most recent auctions. In the latest auction alone, around $6.3bn of these costs were attributed to data centre demand, including forecasts for facilities that have yet to be built.

In effect, households are already paying higher electricity capacity charges, partly because of projected demand from facilities that may never become operational. The Natural Resources Defense Council estimates that, by 2028, the typical household within PJM could face an additional $70 per month in electricity costs if current trends continue. These rising costs are also having political consequences. Recent US polling shows growing public dissatisfaction with data centres, with electricity bills among the main concerns. In response, Virginia has established a separate electricity rate class for data centres, Ohio has introduced minimum billing requirements, and Illinois is debating legislation.

The aim of the Google agreement is to tackle this objection. Its underlying principle is that major new electricity consumers should bring additional generation capacity rather than compete with existing users for an already constrained supply. This approach represents a constructive step towards mitigating the pressure that data centres place on electricity systems.

However, the arrangement remains voluntary. In October 2026, US federal regulators rejected most of PJM’s proposal to require large new electricity consumers to provide their own generation capacity. As a consequence, the industry is voluntarily offering a commitment that regulators have so far declined to impose through PJM’s proposed rules. This poses the question: how long will such voluntary commitments remain viable if the costs of fulfilling them begin to rise?

What the Gulf should take from this

For Gulf countries preparing their own AI infrastructure expansion, the US experience offers a warning about the costs of getting these arrangements wrong.

For countries with abundant energy resources, the expansion of AI infrastructure presents a significant opportunity.

Qatar’s sovereign fund has committed to a $20bn joint venture with Brookfield for AI infrastructure, beginning with a first phase of roughly 300 megawatts and potentially expanding to near 1.5 gigawatts. Abu Dhabi is pursuing plans for up to five gigawatts of AI data centre capacity. Saudi Arabia is targeting almost two gigawatts through HUMAIN. The region’s competitive advantage is substantial. As demand for AI computing grows, access to electricity is becoming an increasingly important factor in determining where computing infrastructure is built. The Gulf is well positioned to capitalise on this demand.

However, the scale of the proposed developments warrants scrutiny. If the proposed 1.5-gigawatt build-out is completed in Qatar, it would represent electricity demand equivalent to approximately 12 percent of the country’s installed generating capacity. Electricity allocated to AI infrastructure cannot simultaneously serve other purposes, meaning the opportunity costs of these projects deserve serious consideration.

In the United States, the debate over AI infrastructure is becoming increasingly centred on who ultimately bears the cost of its electricity demands, with many of the major commitments already in place. The Gulf is in a rare position of being able to settle that question first: what the industrial tariff is, who funds the new generation, what happens to the contracts when the hardware ages, and what long-term economic benefits the country would see beyond energy revenues and a limited number of jobs.

While hosting AI infrastructure can generate considerable economic value, the greater strategic opportunity lies in determining the conditions under which that infrastructure operates. For Gulf countries, the opportunity to establish these terms exists now, before large-scale commitments constrain their options.

The views expressed in this article are the author’s own and do not necessarily reflect Al Jazeera’s editorial stance.

The US tech giant is investing $15bn in a huge new data centre project in northern Finland.

Finnish police have launched a criminal investigation into suspected breaches of environmental law at the construction site of a planned Google data centre in northern Finland.

The project is part of a $15bn deal announced last month by Alphabet’s Google to invest in artificial intelligence (AI) infrastructure in Finland, which the US company described as its largest investment in Europe.

Authorities have said Tuike Finland, a local subsidiary of Google, conducted forest clearing and drainage at a planned site near the town of Muhos, before the environmental impact assessments required under Finnish law had been completed.

“We are investigating whether the company’s operations have breached the Water Act and resulted in environmental damage,” Detective Superintendent Kari Pakanen told the Reuters news agency on Thursday, adding that a guilty finding could lead to a corporate fine.

Pakanen declined to say whether any individuals could become a target of the investigation, and a Google spokesperson said the company would cooperate with the police investigation and continue to focus on maintaining compliance at its sites.

Tree felling prompts local anger

Videos and photographs showing the scale of the forest clearance at the site of the data centre, and Google’s initial reluctance to address the issue in the days after the news broke, have prompted widespread anger in Finland, where there is a high level of respect for the natural environment among the public.

“All the outrage and dismay directed at Google, I can very much share that, especially when we are talking about the parent company valued as the third-largest in the world,” Finnish Green MP Oras Tynkkynen told Al Jazeera. “There is just no excuse. You have to know the legislation of the country you are operating in; you can afford to hire the best legal experts to help you navigate the local legal requirements.”

Finnish authorities ordered Google to stop work at the data centre site at the start of October, and the latest information from the police highlights shortcomings in the drainage and water management at the site.

Tynkkynen called Google’s actions “unforgivable”, but cautioned that while the anger being directed at the US tech giant was justified, it was also diverting scrutiny away from legacy Finnish forestry companies, which he says are “by far the biggest contributors to biodiversity loss in this country”.

What is Finland’s appeal for data centre construction?

Finland’s cool climate, low-carbon electricity and government endorsement of AI projects have contributed to making the country one of Europe’s main destinations for data centre investments in recent years.

Finland, similar in size to Germany – but with one-seventh of the population, less than six million people – has had less vocal opposition to data centre buildouts than other places, including in the United States.

AI data centres are among the main drivers of a projected rise of up to 40 percent in Finland’s electricity consumption in the coming years, if all agreed grid connection projects go ahead, grid operator Fingrid has said.

Google’s planned investments in Finland include data centres, electricity grid improvements and clean energy and battery projects to power services such as Gemini, Search, Maps and YouTube.

Supporting Google’s plan, Finland’s Prime Minister Petteri Orpo said in September that the deal would be a boost for the economy, playing down concerns that it would suck up energy and push up power prices.

As Luiz Inacio Lula da Silva celebrated his election victory on Sunday in Sao Paulo, he placed his hand on the head of a man standing beside him and beaming in a green and yellow Brazilian football jersey.

Lula thanked him for his role in the campaign, calling him “our general”.

Andre Janones, a 42-year-old Brazilian congressman who had spent the final weeks of the presidential run-off building a mass online movement to get Lula re-elected, then embraced the president.

“He played an exponential role in this campaign in lifting the morale of our troops every day, morning, noon and night,” Lula told thousands of supporters who packed Paulista Avenue, the city’s main artery.

Just three weeks earlier, the president seemed to be in trouble. He had finished behind his right-wing rival Flavio Bolsonaro after the first round of voting, his campaign struggling to keep up with the right’s powerful social media operation.

Janones, who had also helped coordinate Lula’s digital operation in his run-off against Jair Bolsonaro in 2022, was convinced Lula’s campaign needed a change – and that he could help turn things around.

He posted a video declaring, “We are going to win this election. Lula will be re-elected president of Brazil.”

His plan centred on a WhatsApp channel he had launched months earlier.

Called Exercito do Lula, or “Lula’s Army”, it had about 60,000 followers after the first round.

By election day, Lula’s Army had grown to almost 11 million subscribers, with Janones widely credited with helping Lula stage his comeback.

How Andre Janones built Lula’s WhatsApp army

Janones is a lawyer and congressman who grew up in a working-class family in Minas Gerais. He made a name for himself through his social media commentary during Brazil’s 2018 truckers’ strike and was elected to Congress that year.

In 2022, he dropped his own presidential bid to endorse Lula, joining his run-off campaign.

“At the time, he became widely known for employing tactics similar to those used by Bolsonarism, despite operating on the left,” said Viktor Chagas, a professor at Fluminense Federal University who studies digital politics and online campaigning.

Janones coined the term “cultural Janonism” to describe his approach, which relied on urgent messages, memes and calls for supporters to take action, as well as what Chagas described as the occasional use of negative campaigning and “content of questionable origin”.

But the scale of Lula’s Army in 2026 was unprecedented, Chagas said.

In the video declaring that Lula would win, Janones urged supporters to enlist as “soldiers” by posting saluting emojis and joining his WhatsApp channel, where he assigned them what he called “missions”.

The tasks were simple, from recruiting three friends to sharing campaign videos to promoting hashtags.

They came in a relentless stream of messages from Janones himself, often written in capital letters. Supporters couldn’t reply, but they could forward the messages or react with emojis, often doing so in the tens of thousands.

Chagas described the strategy as a form of “electoral gamification”, with Janones setting targets for supporters to reach together.

For Andre Pagliarini, an assistant professor of history and international studies at Louisiana State University and the author of a book on Lula, the appeal of Janones’s approach was giving people something they could actually do to help Lula win.

“It became an organising platform that grew exponentially because it spoke to a desire that many Lula supporters evidently had to do something tangible to defeat Flavio Bolsonaro,” he said.

How Lula’s campaign changed its message

But the path to victory wasn’t just about mobilising supporters online.

Pagliarini credited the campaign for also finding a more effective way to tell the story of the president’s working-class past.

“The campaign’s mistake in the first round was to present this history as a finished monument, worthy of admiration but distant from the lives of the people it was meant to represent,” he wrote.

After Bolsonaro mocked the president for having nine fingers during the final televised debate, for example, Lula responded by recalling how he lost a finger at 17 while working a night shift at a factory. He then linked his experience to workers injured as a result of employers’ negligence.

“Unlike Flavio, he reminded voters, he knew the world of work because he had lived in it,” said Pagliarini.

After the debate, Janones spread the message online, promoting the hashtag #SomosTodos9Dedos, or “We Are All Nine Fingers”. Supporters shared the slogan, with some posting photographs of their own amputated fingers in solidarity.

Janones had also publicly called for simpler, less polished campaign videos that showed a more relatable side of Lula, drinking coffee, visiting people’s homes and talking to ordinary Brazilians. Around the same time, the campaign started posting more informal, selfie-style videos.

“Lula was looser in his political communication, less mediated,” Pagliarini said.

Did Lula’s WhatsApp army help him win?

It is difficult to establish whether Janones’s digital army actually delivered votes.

“I am not sure there is a way to empirically connect the group to the outcome,” Pagliarini said.

But he believes Janones helped turn the campaign around.

“Janones himself became a kind of character in the campaign, a cheerleading strategist who seemed confident in a plan that had been missing throughout much of the race,” wrote Pagliarini.

Chagas agreed that it was “difficult to imagine the outcome we achieved without Lula’s Army”, adding that the channel brought together left-wing supporters who had previously been scattered across fragmented online communities, and gave them a way to work together.

The challenge now is keeping that momentum going, Chagas said, pointing to the far right’s use of “permanent campaigning”, maintaining the engagement of supporters between elections.

On election night, at least, Janones kept his soldiers busy.

“EVERYONE TO PAULISTA TO CELEBRATE,” he wrote, urging them to gather on Sao Paulo’s Paulista Avenue. “NOW.”

Later, he asked them not to go to sleep, adding that he’d let them go once he got back from the celebrations.

“Then we’ll discuss a few days off for you,” he joked in a message that drew more than 225,000 reactions.

OpenAI’s announcement is the latest in a series of claimed scientific breakthroughs credited to AI.

OpenAI has announced that its AI models have made several scientific discoveries in mathematics, computer science and physics.

The AI giant said on Monday that the three new discoveries concerned the fields of harmonic analysis, complexity theory and quantum field theory.

Among the discoveries, AI models proved the local smoothing conjecture, one of the central propositions of harmonic analysis, in every spatial dimension for the first time, according to OpenAI.

Researchers in harmonic analysis, a branch of mathematics that studies how complex signals and functions can be broken down into elementary waves, had previously only proved the theory in three dimensions.

‘Historic development’

Alex Iosevich, a professor of mathematics at the University of Rochester who specialises in harmonic analysis, said OpenAI appeared to have “cracked” the conjecture in a “historic development”.

“The conjecture was formulated by my PhD adviser, Chris Sogge, so it is near and dear to my heart, and the news was a shock at first,” Iosevich told Al Jazeera.

“Now I am filled with joy and optimism, because the scientific enterprise as a whole is going to make giant leaps as a result of the AI revolution, and we have the chance to be a part of this historic development.”

Iosevich said he expected the field of theoretical mathematics to “change drastically” due to AI.

“I feel a twinge of sadness as a result,” Iosevich said.

“Many theoretical mathematicians I know feel the same way, but I firmly believe that these feelings will soon give way to the realisation that there is so much we can now accomplish.”

“Problems in pure mathematics did not arise in a vacuum,” he added. “Many of them came out of concrete physical considerations, and these connections can now be explored rapidly and decisively.”

According to OpenAI’s announcement, its models also disproved the Exponential Time Hypothesis, which relates to the time it takes to solve complex computational problems, and provided “mathematically rigorous [proof of the] existence” of quantum chromodynamics, which studies the interaction of quarks and gluons.

The ChatGPT creator said it had reached out to scientists outside the company to turn its “raw research data into scientific understanding”.

The San Francisco-based company said it decided on several different methods for releasing the findings based on consultations with three expert groups chosen to act as “shepherds”.

While the groups collaborating on the mathematics and computer science problems opted to make the AI-produced data public immediately, the physics group decided to defer the release in order to write up their own interpretation of the findings, OpenAI said.

“All of these approaches involve tradeoffs,” OpenAI said.

“On the one hand, the collaborations empower experts to foster unpacking the results and contextualizing them in the broader academic community,” the $852bn start-up said.

“On the other hand, choosing initial shepherds necessarily shapes who gets involved first and creates asymmetry of information. We welcome feedback on how to broaden participation and make this process more useful.”

OpenAI’s announcement is the latest in a series of claimed scientific breakthroughs credited to AI.

Last month, Claude creator Anthropic announced the discovery of a new enzyme said to resemble CRISPR, the groundbreaking gene editing mechanism used in treating conditions such as cancer and sickle cell disease.

Anthropic’s announcement drew mixed reactions among scientists.

While some researchers hailed the finding as an exciting development, others suggested that the AI company had exaggerated its significance.

WASHINGTON — In September, just weeks before Americans were set to vote in this year’s midterm elections, a sudden panic over the dangers posed by artificial intelligence caught the White House off guard. A whistleblower had gained viral attention warning the technology could soon pose an existential threat.

President Trump was quick to dismiss concerns, evoking urgency in the U.S. race to achieve global dominance in superintelligence. He rejected calls to negotiate guardrails on the most advanced AI models with China’s leader, Xi Jinping, weeks later, during a prime opportunity to do so in Washington.

Privately, the president expressed concern that public hysteria over AI could jolt markets that have become excessively concentrated in AI-related stocks. He had already seen fresh polling that showed widespread bipartisan opposition to the boom in construction of AI data centers, a push by Trump that has cannibalized chips and energy from other sectors of the economy, driving up costs.

Now, on the campaign trail in the final days before the vote, Trump is vowing to expand a “golden age” of prosperity through the end of his presidency. But experts say his policies on AI have set the U.S. economy up for a period of tremendous risk over the next two years.

Some of the world’s leading financiers are sounding increasingly alarmed about an AI bubble. Ray Dalio, founder of Bridgewater Associates, has warned that AI stocks have become detached from financial fundamentals, pushing markets to precarious highs. Ken Griffin, the world’s richest hedge fund owner, said last week that AI is “almost certainly” fueling a market bubble that could be on the verge of a burst.

Economists warn that Trump has overseen fiscal policy that would give the federal government fewer tools than it has had in the past to respond to a crisis.

“Trump could easily run into a financial crisis before his term is up,” said Kenneth Rogoff, an economist at Harvard. “It is entirely plausible to have a 30% — or more — drop in the stock market, and in the context of an economy so dependent on AI, this in turn could easily lead to a recession and widespread financial distress.”

Spending on data center infrastructure last winter and spring accounted for nearly all U.S. economic growth, according to independent analyses. Without it, growth would have slowed to a mere fraction of a percentage point, leaving the economy nearly stagnant and flirting with recession.

The Nasdaq 100 has about 70% of its index weight in companies that depend on AI, from semiconductor manufacturers and Big Tech giants to cloud platforms and software providers. Anthropic, one of the top AI industry leaders, plans an initial public offering next month, while its chief rival, OpenAI, expects to go public next year.

From left, Chris Painter, president of the Model Evaluation and Threat Research, Marius Hobbhahn, CEO of Apollo Research, Paul Ohm, professor of law at the Georgetown University Law Center, Kurt Gaudette, senior vice president of Dragos, and Daniel Kokotajlo, executive director of the AI Futures Project, are sworn in to the Senate Homeland Security and Governmental Affairs Subcommittee on Disaster Management, District of Columbia, and Census hearing titled “Rogue AI: Securing the Homeland Against AI Agent Attacks,” in Dirksen Senate Office Building on Wednesday, September 30, 2026.

(Tom Williams/CQ-Roll Call, Inc via Getty Images)

“Because government finances are already strained, with the national debt on a patently unsustainable trajectory, there is no guarantee of a sharp drop in long-term borrowing rates for the government,” Rogoff continued. “So the options for bailing out the financial sector, or fighting recession, could be much more constrained than during the global financial crisis.”

While doubling down on the promise of an untested, frontier technology, Trump’s other economic policies may be hobbling its growth. Tariffs have raised the costs of imported chips, servers, networking equipment, metals and electrical components, and visa restrictions have tied the hands of industry leaders trying to secure the world’s best talent. Runaway federal deficits have driven long-term bond interest rates, adding pressure on AI firms already strapped with debt.

Servaas Storm, an economist closely following the AI sector at the Delft University of Technology, said the industry is already facing severe financial headwinds that could precipitate a panic.

Prices for energy, chips and minerals are driving up the cost of doing business, while early AI adopters such as Microsoft, Meta and Walmart are scaling back its use as employees spend too much time correcting errors, cleaning up low-quality output and addressing security lapses.

Slower adoption could further complicate the search for revenue, particularly for companies already struggling to monetize the technology. For now, hyperscalers are shouldering the cost of financing frontier AI labs that, as Storm put it, “do not have a credible path to profitability.”

“There is indeed an AI bubble in the stock market and in data center construction, which could pop rather soon, as rising debt loads and elevated interest rates could prove too much to sustain it,” Storm said. “That moment is not imminent, but it is approaching — and it will very likely happen before the end of Trump’s presidency.”

An economy already under duress from Trump’s other economic policies — from tariffs and tax cuts to the Iran war — could precipitate further pain, said Darrell Duffie, a finance professor at the Stanford Graduate School of Business.

Leading AI companies are already running “day-after” drills in anticipation of an AI-related crisis within the next year, according to Axios.

Americans head into election day deeply pessimistic about the economy, with 77% rating it fair or poor, according to a Fox News poll. A separate CNBC survey found that just 34% approve of Trump’s handling of the economy, while 35% of Republicans disapprove — a record high within his own party.

Inflation has fallen sharply from its 9% peak in 2022, during a pandemic-era surge, reaching 2.9% by the end of Biden’s presidency. Under Trump, it has climbed once again to 3.7%, driven by rising energy, transportation and housing costs — among the top concerns for voters entering next month’s vote.

Persistent inflation has forced the Federal Reserve to keep interest rates elevated, squeezing borrowers. And those rates — paired with high government spending and concerns over mounting deficits — have pressured the bond market, pushing 10-year Treasury yields to levels not seen since the early 2000s.

“Financial crises, by their nature, are virtually impossible to predict,” Duffie said, asked whether he shared concerns of a potential AI bubble.

“If we know we are going to be in a crisis soon,” he added, “markets advance the crisis to now.”

Tech-heavy Nasdaq falls 2.6 percent as debate rages over whether AI is a bubble.

US tech stocks have dropped sharply amid market jitters over the sustainability of companies’ enormous capital investments in AI.

The Nasdaq Composite, which is dominated by tech firms linked to the AI boom, fell 2.6 percent on Thursday, its steepest fall since the index tumbled 4.2 percent on June 5.

The broader S&P 500 slid 1.21 percent, the benchmark’s biggest drop since a 1.52 percent decline on July 29.

Apart from Apple, each of the “Magnificent 7”, which make up more than one-third of the overall market, saw substantial losses.

Meta dropped nearly 7 percent after it raised its projected capital spending for this year to $135bn to $145bn in its latest earnings report, even as the Facebook parent company’s net profit fell short of expectations.

Amazon fell 4.6 percent, while Nvidia dropped 4.4 percent.

Tesla and Alphabet both fell 3.1 percent, while Microsoft declined 2.2 percent.

Apple rose 1.2 percent.

The sell-off extended to other major tech players, including Intel and Advanced Micro Devices, both of which fell about 8 percent, and Broadcom, which slid 4.3 percent.

Rising oil prices since February, due to the fallout of the US-Israel war on Iran, added to the downbeat mood.

Brent crude futures rose 4 percent on Thursday and continued to climb on Friday morning, taking the international benchmark above $114.20 a barrel, its highest since May.

Stocks slip in Asian trading

Major stock markets in Asia opened lower on Friday, following the losses on Wall Street.

Japan’s benchmark Nikkei 225 fell more than 2 percent in morning trading before recovering some ground, while South Korea’s KOSPI dropped as much as 3.5 percent.

The Hang Seng Index in Hong Kong fell about 0.5 percent.

Jay Goldberg, a senior analyst at Seaport Research Partners in San Francisco, said the swings in the market appeared to be driven more by sentiment than underlying fundamentals.

“In short, nothing has changed. Investors have just decided to be afraid this week. Everyone I speak to points to a different cause for the downfall … It’s just a giant wall of worry,” Goldberg told Al Jazeera.

“There are legitimate concerns about the AI trade, but this level of shift in the markets is driven much more by investor sentiment than any tangible reason.”

US tech companies’ huge investments in AI have stoked intense debate about whether a market bubble is forming similar to the one that preceded the 2000 dot-com crash.

While some Wall Street analysts view current levels of investment in datacentres and graphics processing units as unsustainable, others argue that AI will revolutionise society to a degree that justifies the eye-watering sums involved.

Global AI investment is forecast to reach $2.7 trillion in 2026 and $3.6 trillion in 2027, according to technology insights company Gartner.

Publisher Boreal says investigations confirmed the Hatian-Canadian Thelyson Orelien used AI in his breakthrough novel.

The publisher of best-selling Haitian-Canadian writer Thelyson Orelien says it has severed its professional relationship with the author after confirming he used artificial intelligence to help write his novel, “It Was Either That or Die”.

Boreal said on Wednesday that it conducted a further investigation after allegations surfaced in September to determine whether AI was used to write the novel.

“New information, confirmed over the past few days, indicates that artificial intelligence was used in the writing process” of the novel, a Boreal statement said.

Orelien had previously denied allegations surrounding his breakthrough novel, and last week, the Canadian publisher said it was standing by the 38-year-old writer.

Boreal said it did not know the extent of AI’s use in the writing process “and we will likely never have a definitive answer”.

It said it gathered all the working versions of the novel and hired an outside party to analyse the various drafts.

“The results obtained this week revealed traces of AI tags in the metadata of certain documents,” Boreal said.

The publisher said it had tried to work with Orelien to gather the evidence necessary to prove his innocence.

That effort was inconclusive, and Orelien “subsequently announced that he would not provide any further information,” the publisher said.

‘Slander and heresy’

Orelien’s novel about migration, known by its original French title “C’etait ca ou mourir”, had been scheduled for translation in more than 20 countries.

France’s top book prize, the Goncourt, removed “C’etait ca ou mourir” from its shortlist after tests with the AI detector Pangram indicated a 100 percent probability that it contained computer-generated text.

But Orelien called the allegations “slander and heresy” in a statement sent to AFP earlier this month.

“Here and there I did use tools for research, particularly for foreign idioms,” he said.

“But not a single sentence in this book was created by a machine, nor the meaning or the structure of a story that comes from my gut,” Orelien said.

On Wednesday, he stood firm, telling the French newspaper Liberation that he had used AI “in the creative process” for “historical research”.

The author said he began writing the book in late 2017, before AI services such as ChatGPT were widely available.

When the allegations first surfaced in September, Orelien said he suspected racism was behind the attacks on him.