Russia and Europe on the Brink of a Major War

The Week That Shaped the World — 28 August–4 September 2026

Russia and Europe Edge Closer to War — While Politics, AI and Markets Keep Rewriting the Rules

Some weeks are defined by one event.

This one was defined by the disappearance of boundaries.

Europe still insists it is not at war with Russia, yet its factories arm Ukraine, its airspace increasingly intersects with the conflict, and its governments now speak openly about sabotage, deterrence and military readiness. Moscow, meanwhile, argues that the distinction between support and participation is becoming meaningless.

Elsewhere, the same pattern appeared in quieter forms.

Reform UK discovered that outsider politics eventually meets the machinery of political finance. India produced growth numbers strong enough to make economists suspicious. Uber cut thousands of jobs while investing billions in a future with fewer drivers. New York decided that children should learn to think before machines learn to think for them.

China launched another serious challenge to America’s dominance in reusable rockets, while investors prepared to value a young AI company at tens of billions before its future had fully arrived.

The thread connecting all of this is simple.

Systems built on old assumptions are beginning to behave differently.

And when the rules change faster than the institutions built to manage them, risk stops looking exceptional.

It starts looking normal.

“The most dangerous moment is not when the old order collapses. It is when everyone still behaves as if it hasn’t.”

1. Russia and Europe on the Brink of a Major War

Europe and Russia are still at peace.

At least on paper.

Germany is now investigating a growing number of incidents involving critical infrastructure, drones and suspected sabotage. Berlin has accused Russia of involvement in the attempted drone attack near Leipzig/Halle Airport. Moscow rejects the accusation and says Europe is deliberately constructing a narrative of Russian aggression.

Not every incident has been attributed to Russia. That matters. A damaged power line is not automatically an act of Moscow, and suspicion is not evidence.

But the pattern around the conflict is becoming harder to ignore.

Europe formally insists it is not a party to the war in Ukraine. Yet much of Ukraine’s military rear now sits inside Europe. Weapons are manufactured in European factories. Ammunition is financed by European governments. Ukrainian troops are trained on European soil. Long-range systems supplied by Britain, France and others are used against Russian forces and, increasingly, against targets inside Russia.

From Europe’s perspective, this is legitimate support for a country defending itself from invasion.

Russia sees the same arrangement differently.

Moscow increasingly argues that states supplying weapons, intelligence, training and industrial capacity are no longer politically neutral, even if they remain legally outside the conflict. On 27 August, Russian officials warned that British military assets inside and outside Ukraine could become potential targets in response to strikes on Russian territory using British-supplied weapons. Britain rejected the threat and reaffirmed its support for Kyiv.

The Baltic region adds another uncomfortable layer.

This is not a development of the past few days. It is useful precisely because it shows how long the boundary has already been eroding.

During Ukrainian strike operations against targets around Russia’s Baltic coast, Ukrainian drones have entered Estonian airspace. Tallinn has said these were not authorised attack corridors and, in some cases, linked the deviations to Russian electronic warfare.

Legally, that distinction is important.

Strategically, it is becoming less comforting.

Europe is supplying the war from behind NATO borders while Russia increasingly questions whether those borders can indefinitely shield the military infrastructure supporting Ukraine. At the same time, European governments are expanding defence budgets, increasing weapons production and openly preparing their societies for a more dangerous confrontation with Moscow.

Russia, meanwhile, describes this rearmament as evidence that the West is preparing for a direct conflict.

There is no public evidence that either Russia or Europe has made a decision to begin a major war.

That may no longer be the most important question.

Wars sometimes begin because both sides believe they are responding defensively to the previous move. One side arms because it feels threatened. The other sees the arms as confirmation of hostile intent and responds. Each new step then becomes justification for the next.

If Europe’s current political course towards deeper military integration with Ukraine and confrontation with Moscow does not materially change — and if Russia continues answering it with military pressure, threats and hybrid operations — the possibility of a direct Russia–Europe conflict is becoming increasingly difficult to dismiss.

The danger is not necessarily that somebody wants a European war.

It is that the system is beginning to produce one.

“Wars are not always chosen. Sometimes both sides simply keep closing the exits.”

2. Reform UK Discovers That Outsiders Need Accountants Too

Reform UK built much of its appeal on a simple proposition: Westminster had become a closed club, and Nigel Farage intended to kick the door open.

This week the paperwork followed him through it.

Two senior Reform figures, James Orr and Dan Jukes, stepped down while the party investigates allegations arising from an undercover Channel 4 investigation into political donations. Reporters posing as American donors discussed funding the party, including polling reportedly worth around £30,000.

The allegation is serious because British law restricts political donations from impermissible foreign sources.

The undercover footage also reportedly captured discussion of whether money from an American donor could reach Reform through his UK-based son. That does not establish that an unlawful donation was ultimately made. Reform says it is investigating, while Farage insists the party has broken no laws. Labour has asked the Metropolitan Police to examine the matter.

That distinction matters.

An embarrassing conversation is not a conviction.

But Reform has a political problem even before lawyers decide whether it has a legal one.

Insurgent parties sell moral distance from the establishment. Their great advantage is being able to point at Westminster’s habits — donors, lobbying, favours, opaque networks — and say: we are not them.

The moment questions about money arrive at their own door, that distance shrinks rather quickly.

There is another irony here.

Reform wants to present itself as the defender of British sovereignty. An argument over whether foreign money might have found a route into British politics is therefore particularly awkward territory.

The established parties will enjoy the spectacle.

They should perhaps resist enjoying it too much. Britain’s political financing system has hardly earned the right to behave like a monastery discovering gambling in the village.

For now, these remain allegations.

But Reform has discovered one of politics’ less glamorous laws: rebellion may get you into Westminster. Once there, the Electoral Commission still wants the receipts.

“The outsider’s greatest political asset is distance from the system. Money has a habit of shortening the journey.”

3. Germany and Russia Begin Closing the Last Doors

Wars usually destroy bridges.

This confrontation has started with language schools.

Russia announced this week that it would close Germany’s Goethe-Institut operations in the country after Berlin moved against the Russian House cultural centre and Russian diplomatic facilities in Germany following the Leipzig/Halle drone dispute. Moscow describes its decision as retaliation. Germany says its measures followed what it considers Russian involvement in hostile activity on German soil. Russia denies responsibility.

The Goethe-Institut survived something rather remarkable.

It continued teaching German, running libraries and maintaining cultural links through years in which almost every other part of the German-Russian relationship was collapsing.

Its president, Gesche Joost, called the closure “the end of an era” in cultural relations between the two countries. The institutes that emerged after the Cold War were intended to connect societies precisely when governments disagreed.

Now disagreement has swallowed the connection.

This will not alter the balance of power in Ukraine.

No tank will stop because a language class has been cancelled.

That is precisely why the story matters.

Cultural institutions belong to the infrastructure governments normally preserve when they expect relations eventually to recover. Embassies shrink. Trade falls. Sanctions come and go. Somewhere, quietly, students still learn each other’s language.

When even that becomes politically unacceptable, the assumption begins to change.

Perhaps there will be nothing to return to.

Germany sees Russian hybrid operations as an expanding security threat. Moscow says Berlin is deliberately dismantling bilateral relations while supporting Ukraine militarily.

Both governments therefore increasingly describe the other not as a difficult partner, but as a hostile actor.

Diplomacy rarely dies in one dramatic moment.

More often the lights simply go out, room by room.

“Closing a cultural institute will not start a war. It merely removes one of the places where people might someday remember how to end one.”

4. Kyiv Learns What “Another Night” Means

Kyiv has developed an unpleasant new unit of time.

Another night.

Another air-raid warning. Another wave of drones. Another search through damaged apartments when daylight returns.

Russian forces have subjected the Ukrainian capital and surrounding region to more than a week of repeated air attacks. On one of the deadliest days, 12 people were reported killed in the Kyiv region, including eight railway workers. A later attack injured ten people, among them a child and six medical workers, and set fire to a high-rise residential building.

Moscow says its strikes target military and military-linked infrastructure and denies deliberately targeting civilians.

Kyiv says Russia is using large-scale drone and missile attacks to terrorise the population and exhaust Ukraine’s air defences.

Both claims must be separated from what can be independently established.

Military and logistical infrastructure does sit inside functioning cities.

So do civilians.

The more important military story may be accumulation.

One raid can be intercepted.

Repeated raids become an arithmetic problem.

Every Patriot interceptor fired today is one unavailable tomorrow. Every radar crew works another shift. Every repair team returns to infrastructure that may be struck again.

Ukraine has been asking allies for additional air-defence systems and missiles precisely because Russia appears increasingly able to sustain this pressure.

Ukraine, meanwhile, continues striking Russian energy and logistics infrastructure of its own, including facilities around the Baltic region.

The war therefore increasingly operates as two competing attempts to make the rear less comfortable than the front.

Factories, railways, ports, power systems and cities become part of the military equation.

Politicians can discuss negotiations while this happens.

People beneath the drones experience diplomacy somewhat differently.

For them, a ceasefire is not a communique.

It is silence.

“Diplomats measure progress in meetings. A city under drones measures it by whether the children sleep through the night.”

5. Trump Sends Europe the Bill for Ukraine

Europe has spent four years discussing the price of defending Ukraine.

Donald Trump would now like to discuss the invoice.

The US president said this week that Washington intends to seek repayment from European countries for American weapons and military assistance previously supplied to Ukraine. He criticised the Biden administration for providing equipment without demanding sufficient payment from Europe.

There is, so far, no detailed public mechanism explaining exactly how much Trump intends to recover, from whom, or under what legal arrangement.

That rather important accounting exercise has been left for later.

But the political message is considerably clearer.

America no longer wants the European security order to operate on the assumption that Washington supplies the strategic depth while Europe supplies statements of solidarity.

This transition had already begun.

Under the NATO Prioritised Ukraine Requirements List, European allies finance packages of American weapons for Ukraine. Trump is now pushing the logic backwards as well: if previous American administrations carried too much of the cost, Europe should compensate them.

From Washington’s perspective, there is an obvious argument.

Ukraine is geographically European. European governments describe its survival as fundamental to their own security. American stockpiles are finite, and the United States now faces military demands elsewhere, including the confrontation with Iran.

Why, then, should the American taxpayer permanently underwrite the largest security crisis on Europe’s own continent?

Europe has answers.

The United States benefits from NATO. American defence companies profit from European orders. Washington gains enormous strategic influence from leading the alliance.

All true.

None of them make the weapons free.

And this is where the conversation becomes uncomfortable for Europe.

For years Brussels spoke enthusiastically about “strategic autonomy.” Washington now appears increasingly willing to test whether Europe meant it.

Autonomy sounds rather elegant in a summit declaration.

It becomes less romantic when somebody attaches a price.

“Europe spent years asking for strategic autonomy. Washington has begun asking whether it would like the receipt.”

6. India’s Economy Is Growing So Fast That Economists Are Checking the Calculator

Most governments dislike disappointing GDP numbers.

India has acquired the opposite problem.

The economy expanded by 7.8% year-on-year in the April-to-June quarter, significantly above economists’ expectations of around 7.1%. For a major economy, that is an extraordinary pace.

Ordinarily, governments would frame the number, investors would applaud and everybody would proceed to the next quarter.

Instead, prominent Indian economists have started asking whether the figure is quite as strong as it looks.

Former Reserve Bank of India governor Raghuram Rajan and former finance ministry official Subhash Chandra Garg are among those questioning elements of the methodology. India revised its GDP framework earlier this year, changing its base year, expanding data sources and altering the way inflation is stripped from nominal growth.

That matters because real GDP is partly a calculation of what remains after price increases are removed.

Underestimate inflation and growth appears larger.

The Indian government rejects suggestions that the statistics are misleading and argues that the revised methodology provides a more accurate picture of a rapidly changing economy.

And there is genuine evidence of strength.

Car sales and bank lending support the story of robust domestic activity. Investment has not collapsed. India continues attracting companies looking for manufacturing alternatives to China.

Other indicators are less enthusiastic.

Employment creation remains a persistent weakness. Foreign investment has not risen in proportion to the headline growth rate, while some purchasing-manager data have softened.

None of this proves that India’s boom is fictional.

Quite the opposite.

The interesting question is whether a genuinely fast-growing economy is being made to look even faster by statistical architecture.

India may indeed be becoming one of the great economic stories of this decade.

Strong stories still deserve good arithmetic.

“When GDP disappoints, governments question the economy. When GDP becomes spectacular, economists start questioning GDP.”

7. Uber Is Preparing for a World That Needs Fewer Uber Drivers — and Fewer Uber Employees

Uber is cutting approximately 3,300 jobs.

About 10% of its workforce.

The company says the restructuring will simplify management, reduce layers and allow decisions to be made faster. It is Uber’s largest round of job cuts since the pandemic.

That would already make this a significant corporate story.

Then come the robots.

Uber is preparing to invest more than $10 billion in autonomous-vehicle technology as Waymo, Tesla and others move closer to making driverless taxis commercially viable at scale. Its ambition is to become the marketplace through which passengers order autonomous vehicles even when Uber does not manufacture the cars itself.

This produces a peculiar transition.

A company whose great innovation was organising millions of human drivers through software is now spending heavily on a future in which the driver may disappear.

And before that technology has fully arrived, thousands of corporate employees are disappearing too.

AI is part of the pressure across the technology sector, though Uber says these cuts are primarily about organisational efficiency and the autonomous-vehicle challenge.

The economics are fairly merciless.

Human drivers are expensive, variable and legally complicated.

Autonomous vehicles are ruinously expensive to develop but potentially cheaper to operate once the technology works.

The phrase doing the most work there is once it works.

Robotaxis still face regulatory, safety and geographic limitations. Human drivers remain fundamental to Uber’s present business.

So the company must finance tomorrow while continuing to operate today.

That frequently means asking today’s workers to pay part of tomorrow’s bill.

Markets generally approve of such things.

Employees tend to experience efficiency somewhat more personally.

“The future of transport may be driverless. The road towards it is already losing passengers from the payroll.”

8. New York Tells Children to Learn Before Asking the Machine

New York City has done something unfashionable.

It has told technology to wait.

Beginning this school year, the largest public-school system in the United States will impose a one-year moratorium on student-facing generative AI from pre-kindergarten through eighth grade, affecting almost 600,000 children. High-school students will retain limited, supervised access through AI-literacy programmes and selected pilots.

Mayor Zohran Mamdani’s argument is almost aggressively old-fashioned.

Children need teachers.

They need other children.

And before outsourcing difficult thinking to a machine, they ought to experience some difficult thinking.

Teachers will still be allowed to use AI for tasks such as lesson planning, while older students will study how the technology works, including its biases and risks. The city therefore is not banning knowledge of AI.

It is delaying dependence on it.

Critics see a different danger.

AI will exist whether schools approve of it or not. Wealthier children will continue using the best tools privately, potentially leaving public-school pupils less prepared for universities and workplaces where AI literacy becomes expected.

That is a serious argument.

So is the argument on the other side.

We still know remarkably little about what happens when children routinely outsource writing, problem-solving, memory and interpretation while their cognitive habits are still being formed.

Adults gained AI after learning how to think without it.

Children increasingly face the possibility of learning both at the same time.

That is not necessarily progress.

It is an experiment.

New York has decided that 600,000 children should not automatically be the control group.

The technology industry will dislike the precedent.

Parents elsewhere may find it rather interesting.

“AI can help a child find an answer. New York is asking whether the child should first learn how to have the question.”

9. China Launches Its Falcon 9 Question

For years, SpaceX had a particularly useful competitive advantage.

Almost everybody wanted reusable rockets.

SpaceX actually had them.

China has just moved another step towards changing that equation.

Galactic Energy, a private Chinese aerospace company, successfully completed the maiden orbital flight of its PALLAS-1 rocket this week, placing its payload into the intended orbit. The 52-metre, two-stage liquid-fuel launcher can carry roughly five to seven tonnes into low Earth orbit.

The important word, however, is not launch.

It is reusable.

PALLAS-1 has been designed with a recoverable first stage, including throttleable liquid oxygen-kerosene engines and the systems required for future vertical landing.

There was no recovery attempt on this first flight.

That distinction is important. China has successfully launched a rocket designed for reuse; it has not yet demonstrated the Falcon 9 trick of routinely returning, refurbishing and flying the booster again. Galactic Energy says recovery tests will follow.

SpaceX’s advantage was never simply landing a rocket once.

It was making landing one sufficiently boring to become an industrial process.

That changed launch economics.

Boosters became assets rather than expensive fireworks. Flight cadence rose. Satellite constellations became easier to build. Space infrastructure accelerated.

China understands this perfectly.

Galactic Energy is only one of several Chinese companies pursuing reusable launchers. LandSpace, iSpace and others are working on similar systems as Beijing encourages a larger commercial-space industry alongside its state programme.

Competition in space is therefore moving away from the old question of who can reach orbit.

Several countries can.

The next question is who can reach it cheaply, repeatedly and on Tuesday morning because a customer asked.

PALLAS-1 has not caught SpaceX.

It has shown that China intends to run the same race.

“The reusable rocket revolution belonged to SpaceX when landing boosters looked impossible. China is entering the competition now that they look like an industry.”

10. Thinking Machines and the $40 Billion Price of Expectation

Thinking Machines Lab is reportedly discussing raising another $1 billion.

The proposed valuation is at least $40 billion.

The company was founded only last year.

Welcome to the mature phase of the AI boom.

Accel is reportedly in talks to lead the new financing for the company founded by former OpenAI chief technology officer Mira Murati. Thinking Machines previously raised around $2 billion at a $12 billion valuation, one of the largest seed rounds in technology history.

There is now a business.

The company launched Inkling, an open-weight model, and charges customers for computing used to adapt models to proprietary data through its Tinker platform. TechCrunch reports annualised revenue above $100 million, citing a source familiar with the company.

That makes the valuation easier to understand.

Not necessarily easier to justify.

At $100 million in annualised revenue, a $40 billion valuation implies that investors are paying today for an extraordinary amount of tomorrow.

This is increasingly normal in artificial intelligence.

Capital is not merely chasing current revenue. It is trying to purchase position before the industry settles — research talent, computing access, intellectual property and, perhaps most importantly, the possibility that one laboratory becomes part of the infrastructure on which everybody else eventually depends.

The risks are equally obvious.

Thinking Machines has already seen several prominent departures, including researchers who returned to OpenAI. AI models are expensive to train. Competitors are extraordinarily well funded. Technical advantages can disappear within months.

And yet investors remain willing to assign valuations normally associated with mature multinational businesses to laboratories still defining their products.

Perhaps they are right.

The next Microsoft or Google may indeed be taking shape in a research lab right now.

But there is something revealing about an economy in which the expected value of future intelligence rises faster than the companies have time to generate present revenue.

The AI boom is no longer waiting for profits before deciding what things are worth.

Expectation has become an asset class of its own.

“Thinking Machines may eventually justify $40 billion. Investors have simply decided they would rather pay before finding out.”