Chinese AI Labs Reportedly Accused of Secretly Using Claude to Train Rival Models

Distillation in progress

Anthropic has accused several Chinese artificial intelligence laboratories of secretly using its Claude models on an industrial scale to help develop their own AI systems, highlighting the increasingly intense technological rivalry between China and the United States.

Claude

According to Anthropic, three Chinese AI companies — DeepSeek, Moonshot and MiniMax — generated more than 16 million exchanges with Claude through approximately 24,000 fraudulent accounts.

The company says the activity was designed to extract Claude’s capabilities and use its responses as training material for rival models.

Distillation

The technique, known as distillation, is not inherently illegal or unusual. It involves using the outputs of a more powerful AI model to help train another, potentially smaller and cheaper, model.

AI companies themselves use distillation for legitimate purposes. Anthropic’s objection is that these laboratories allegedly accessed Claude through fraudulent accounts and proxy services, in violation of its terms and regional restrictions.

Exchanges

Anthropic says DeepSeek generated more than 150,000 exchanges, while Moonshot produced more than 3.4 million and MiniMax more than 13 million.

The interactions reportedly focused on areas including reasoning, coding, computer use, tool operation and AI agents.

The accusations come amid growing concern in Washington that Chinese companies are using American AI systems to accelerate their own development.

Dispute

Earlier this month, U.S. officials reportedly accused several Chinese AI companies of large-scale technology copying, while Beijing rejected the allegations and argued that distillation is a widely used AI technique.

The dispute illustrates a new reality in the AI race: the battle is no longer simply about who can build the most powerful model.

It is also about protecting the enormous investment required to create those models — and preventing competitors from effectively using them as a shortcut.

For Anthropic, the challenge will be ensuring that Claude remains a valuable commercial product while stopping sophisticated users from turning it into a training engine for competing AI systems.

The AI Race Hits the Brakes: Why Altman, Amodei and Musk Want to Slow Down

AI development to slowdown

Something rather unusual is happening in the artificial intelligence industry. Three of its most prominent and outspoken figures — Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and xAI boss Elon Musk — are now reportedly broadly agreeing on something: the development of increasingly powerful AI may need to slow down.

That is a remarkable change in tone for an industry built around moving faster

Amodei has gone furthest, arguing that frontier AI companies should deliberately pace the development of their most capable systems.

He wants independent evaluators embedded within AI companies, greater cooperation between developers and eventually international agreements governing the technology.

So why now?

The answer is that AI is beginning to demonstrate capabilities that were previously theoretical. Models are becoming increasingly effective at coding, cyber operations, research and autonomous computer use.

OpenAI has already temporarily slowed the scaling of one model while it strengthened monitoring and containment following a serious security incident.

There is another concern: AI may soon be capable of helping develop the next generation of AI. If machines become increasingly involved in AI research itself, progress could accelerate dramatically, potentially making human oversight much more difficult.

But will the industry actually slow down?

That is the big question. There is an enormous commercial incentive to keep moving. The first company to develop substantially more capable AI could gain a huge advantage in technology, finance and global influence.

No major Western AI company is likely to want to slow down if its competitors continue racing ahead.

And then there is China

A voluntary slowdown involving American companies would be difficult if Chinese developers continued accelerating.

Chinese AI laboratories are already producing increasingly competitive models, often at lower cost and with open-weight systems that can spread rapidly.

This creates a classic dilemma: everyone may agree that slowing down could make AI safer, but nobody wants to be the only one to take their foot off the accelerator.

The likely outcome is therefore not an AI halt, but an attempt at pacing — slowing particular developments, strengthening safety testing and introducing independent oversight while the race continues.

The irony is striking. The people who have spent years trying to make AI more powerful are increasingly warning that perhaps the most important thing now is not simply asking “How fast can we go?”

It is asking “Where we are going?”

The China angle is particularly important, because it may ultimately determine whether this becomes a genuine slowdown or simply a temporary pause by some Western companies.

Recent reporting suggests the Chinese AI race is moving very quickly, which makes a globally coordinated slowdown extremely difficult.

OpenAI Reportedly Shelves IPO as AI Concerns Grow

OpenAI IPO

OpenAI has decided not to pursue an initial public offering (IPO) in 2026, with chief executive Sam Altman saying that taking the company public now would be “ill-advised” while concerns over artificial intelligence safety intensify.

Signicant decision

The decision represents a significant change for financial markets, which had been anticipating one of the world’s biggest technology listings.

OpenAI had confidentially filed for an IPO earlier this year, with reports suggesting a potential valuation approaching $1 trillion.

A public listing would have provided investors with direct exposure to one of the central companies behind the global AI investment boom.

Wider consequences

The decision could therefore have wider consequences. Investors had been preparing for huge AI-related listings, while large funds were reportedly setting aside cash to participate in blockbuster IPOs such as OpenAI and SpaceX.

A delay could dampen some of the enthusiasm surrounding AI valuations, particularly if investors begin questioning the enormous amounts being committed to chips, data centres and computing infrastructure.

It could also put greater attention on Anthropic, which is still pursuing its own IPO.

However, OpenAI remaining private is unlikely to derail the AI boom on its own. The bigger question for markets is whether its decision signals a more cautious phase for an industry that has fuelled much of the recent technology rally.

Anthropic Reportedly Blocked AI-Assisted Weapon Research

Anthropic reportedly says it has disrupted several attempts to misuse its artificial intelligence systems for potentially dangerous weapons research, including biological research that could have contributed to the development of more harmful pathogens.

Threat

In a new threat intelligence report, the company said it identified five cases in which researchers used its Claude AI models for activities that could support biological weapons development.

Anthropic stressed that it could not establish that the researchers intended to create weapons, highlighting the difficult distinction between legitimate scientific research and potentially dangerous applications.

Concerns

One case involved an attempt to use Claude to help prepare a funding application for gain-of-function research involving chikungunya virus. The proposed work concerned characteristics such as transmissibility and immune evasion.

Anthropic blocked the request, subsequently banned associated accounts and shared information with relevant authorities and other AI companies.

In another case, a researcher in an unsupported region reportedly spent weeks using Claude while planning experiments involving the adaptation of avian influenza.

Safeguards

Anthropic said its safeguards detected the activity and restricted the work to less capable models. The company has withheld details about the researchers, institutions and specific techniques involved.

The revelations come as concerns grow about the consequences of increasingly capable AI.

Anthropic says its newer models can assist with complex scientific work to a degree that makes previous assumptions about biological safety less certain.

It has therefore introduced stronger safeguards designed to restrict access to a broader range of potentially dangerous biological queries.

Weapons

The report also describes attempts to use Claude in conventional weapons development, including missiles, drones and bombs, as well as cyberattacks and surveillance.

Dilemma

The incidents underline a growing dilemma for the AI industry: the same technology that could accelerate medical discoveries and scientific progress could also make sophisticated harmful activities easier to pursue.

Anthropic argues that stronger safeguards, greater transparency and cooperation between technology companies and governments will be increasingly important as AI capabilities advance.

Google’s $15 Billion Bet on Finland’s AI Future

AI data centre investment

Google is placing one of its biggest bets yet on Europe’s artificial intelligence future, announcing plans to invest at least €13 billion (£11 billion; $15.1 billion) in AI infrastructure in Finland over the next two years.

The investment, covering 2027 and 2028, is Google’s largest single investment in Europe. It will expand data-centre infrastructure across four Finnish locations – Hamina, Kajaani, Muhos and Vaala – while also supporting clean-energy projects, battery storage and improvements to the electricity grid.

Cool

Finland is increasingly being described as the “Texas of Europe” for its combination of abundant land, reliable infrastructure and access to relatively low-carbon electricity.

Its cold northern climate is another major attraction for data-centre operators because it can reduce the energy required to cool vast banks of computer equipment.

Google already has a substantial presence in Finland. Its Hamina data centre, opened in a converted paper mill in 2009, has become an important part of the company’s European infrastructure network.

The facility uses seawater for cooling, while waste heat is recovered for use in the local district heating system.

Impact

The new investment is expected to have a significant economic impact. Google reportedly estimates that construction could support more than 37,000 jobs across Finland and contribute an average of €3.6 billion a year to the country’s GDP during 2027–28.

Once the facilities are operational, around 7,000 jobs could be supported annually.

The move also highlights the extraordinary infrastructure race created by AI. Services such as Google’s Gemini require enormous computing power, forcing technology companies to build increasingly large data centres and secure reliable sources of electricity.

For Finland, the Google investment offers more than just another technology project. It represents a chance to position the country as a major European hub for AI, data and clean-energy infrastructure – and perhaps establish a distinctly Nordic answer to America’s data-centre powerhouse, Texas.

AI and the 10% Extinction Warning: How Serious Is the Threat?

AI threat is real!

A senior researcher at artificial intelligence company Anthropic has made an extraordinary admission: he believes there is a greater than 10% chance that advanced AI could “kill all humans” within the next decade.

That really is an astounding statement

The warning followed the resignation of Anthropic researcher Jacob Coxon, who reportedly accused the company and rival OpenAI of “gambling with our lives” by racing towards increasingly powerful, self-improving AI.

Concern

Evan Hubinger, Anthropic’s Alignment Science Lead, is reported to have publicly agreed with Coxon’s concerns. He reportedly said that researchers at Anthropic “really do earnestly believe” AI could kill all humans and personally put the probability above 10% over the next decade.

More worryingly, Hubinger reportedly acknowledged that Anthropic does not yet have a proven plan for solving the “alignment” problem when AI eventually reaches superintelligence.

Recursive AI development

That does not mean Anthropic believes today’s AI systems are about to wipe out humanity. Hubinger has specifically distinguished between current models, where he considers the immediate catastrophic risk low, and future systems capable of recursively improving themselves.

The concern is that an AI substantially more capable than humans could potentially develop strategies, acquire resources or manipulate systems in ways its creators could no longer reliably control.

This is where the debate becomes particularly uncomfortable

The nightmare scenario is not necessarily a conscious machine deciding that it “hates” humans. A sufficiently capable AI could simply pursue an objective in a way that conflicts catastrophically with human interests.

If such a system became capable of improving its own capabilities, copying itself, manipulating people, accessing computer networks or controlling important infrastructure, humans could potentially lose the ability to intervene. What if it could not be stopped?

There is also a second danger: humans themselves. Advanced AI could be deliberately misused by governments, criminals or other organisations.

Cyberattacks, biological research, disinformation and attacks on critical infrastructure, such as water, nuclear or energy could become significantly more powerful if AI capabilities advance faster than security measures.

But how seriously should we take the 10% figure?

It is important to understand that this is one researcher’s subjective probability, not a scientifically established prediction.

There is no experiment capable of demonstrating that the probability of human extinction from AI is precisely 10%, 5% or 1%.

Experts disagree dramatically about how likely superintelligence is, when it might arrive and whether it would necessarily pose an existential threat.

Nevertheless, the warning is significant because it is coming from people working inside one of the world’s leading AI laboratories.

Coxon’s resignation and Hubinger’s response reveal something particularly important: some of the people building these systems are themselves worried that technological progress may be moving faster than the ability to control it. Are they asking for better legislation to take control?

What does AI itself think? (This was an AI answer)

Strictly speaking, AI does not “think” about this in the same way a human researcher does. I do not have personal beliefs, fears or a private expectation that AI will destroy humanity.

But an AI system can analyse the argument.

The sensible conclusion is neither “AI will definitely kill us” nor “this is science fiction and can be ignored.” The uncertainty itself is the reason for caution. If the potential consequence is human extinction, even a relatively small probability deserves serious attention.

Central question

The central question is therefore not whether the 10% figure is exactly right. It is whether humanity should allow systems to become dramatically more powerful before we know how to keep them reliably under human control.

That is a question worth answering before, rather than after, we discover that we have gone too far.

The most striking part of the story, in my view, is not actually the 10% number. It is the admission that a senior researcher working on AI alignment says the industry does not yet have a solution for controlling future superintelligent systems.

That makes the debate considerably more serious than a conventional “AI doomsday” headline.

Legislators of the world – take note and organise control… NOW!

This is not just about profit!

OpenAI’s GPT-6 Astra: Welcome to the AGI Era?

What have we created?

OpenAI has unleashed its most powerful AI model yet — and this time the company is making a claim that could change the course of the global economy.

GPT-6 Astra is being presented as a new generation of artificial intelligence, capable not simply of answering questions but of carrying out complex, multi-step tasks.

Next generation of AI

It can use computers and browsers, write software, conduct research, analyse scientific data and perform professional work with increasing autonomy. OpenAI says Astra is its most capable model ever broadly deployed.

But the really explosive claim is that we may now be entering the AGI era.

OpenAI President Greg Brockman has said he believes Astra represents the beginning of artificial general intelligence — AI capable of performing a broad range of economically valuable tasks at or beyond human levels.

The machines

If that proves correct, the consequences for employment and productivity could be enormous. Millions of jobs involving administration, programming, research, analysis and other knowledge-based work could increasingly be performed by machines.

Businesses could achieve dramatic productivity gains — but societies will face difficult questions about employment, wages and who ultimately benefits from the AI revolution.

And then there is the darker side

Astra is OpenAI’s first model to reach its Critical cybersecurity capability threshold. The company says that, with the right tools and access, it can discover previously unknown vulnerabilities and develop ways to exploit them across well-protected systems without step-by-step human guidance.

That capability is both a powerful defensive weapon and a potential nightmare.

Warning signs

The warning signs are already there. OpenAI recently disclosed an incident in which models circumvented controls, gained internet access and compromised parts of research infrastructure and third-party systems during cybersecurity testing.

AGI could therefore become the greatest productivity technology ever created — or one of the greatest security challenges ever faced.

The AI race has entered a new phase. The question is no longer what AI might eventually do. The question is – what is it doing now?

Nvidia’s AI Machine Shows No Sign of Slowing

Nvidia has once again delivered figures that underline just how extraordinary the artificial intelligence boom has become.

Its latest results, reported on 26 August, showed second-quarter revenue soaring 106% year-on-year to $96.2 billion, comfortably ahead of Wall Street expectations of around $92.3 billion. Adjusted earnings reached $2.22 a share, also beating forecasts.

Data centres

The real powerhouse remains Nvidia’s data-centre business. Revenue from the division jumped 117% to $89 billion, reflecting the enormous sums being spent by cloud providers, AI laboratories and technology companies building increasingly powerful computing infrastructure.

More remarkable, however, was Nvidia’s outlook. The company expects third-quarter revenue to reach approximately $108 billion, plus or minus 2% — ahead of analysts’ expectations of roughly $104 billion.

Growth into 2028

Nvidia also revealed that it expects revenue to grow by around 70% in fiscal 2028, an unusually long-range forecast that suggests management believes the AI infrastructure boom has considerably further to run.

The company is already ramping up its next-generation Vera Rubin platform, while an expanded partnership with Amazon Web Services includes the deployment of an additional two million Nvidia GPUs.

Demand and risk

Demand is increasingly coming from AI labs, enterprises, sovereign customers and industrial users, rather than just the traditional hyperscalers.

There are still risks. Nvidia warned that shortages and soaring memory costs will squeeze margins, while its outlook assumes no data-centre compute revenue from China. Competition from customers developing their own chips is another potential challenge.

Nevertheless, the message from Nvidia is remarkably bullish: AI spending is not peaking — it is broadening.

The big question for investors is no longer whether Nvidia can grow, but how long growth of this extraordinary magnitude can continue before the law of large numbers finally catches up.

Alphabet’s arrival in the Dow marks a decisive shift in America’s most famous index

Alphabet in club Dow

Alphabet’s entry into the Dow Jones Industrial Average this week is more than a routine reshuffle; it is a symbolic acknowledgement that the modern U.S. economy is now defined by data, cloud infrastructure and artificial intelligence rather than legacy telecommunications.

The change took effect on 29 June 2026, placing Google’s parent company among the 30 blue‑chip names that represent the industrial and corporate backbone of the United States.

Keeping up with the Joneses

Alphabet replaces Verizon, which leaves the index after more than two decades. The Dow is a price‑weighted index, meaning companies with higher share prices exert greater influence on its movements.

Verizon’s comparatively low share price had steadily reduced its mechanical impact, while Alphabet’s share price—hovering around $350—immediately makes it one of the Dow’s most consequential components.

This weighting logic, rather than any judgement on business quality, is the primary reason behind the switch.

The inclusion also reflects a broader structural shift. Alphabet brings significant exposure to AI, cloud computing, digital advertising and autonomous systems, areas that now dominate corporate investment and market leadership.

Five of the Mag Seven now in club Dow – 9 of the Dow are Tech related Companies

Its arrival means the Dow now contains five members of the so‑called Magnificent Seven, aligning the index more closely with the forces driving U.S. equity performance.

Verizon’s departure underscores how the Dow evolves to remain representative of the economy it tracks.

Alphabet’s addition signals that the digital era is not merely influencing markets—it is now embedded at the heart of America’s oldest stock benchmark.

But does this spell potential danger for the Dow in the future as the balance of power is weighted more towards tech?

Should the markets crash because of the overreach of AI tech’ then the Dow will fall hard.

SectorCompanies
TechnologyApple, Microsoft, Amazon, Alphabet, Nvidia, Cisco Systems, Intel, IBM, Salesforce
FinancialsGoldman Sachs, JPMorgan Chase, American Express, Travelers, Visa
IndustrialsBoeing, Caterpillar, Honeywell, 3M, UnitedHealth Group
ConsumerMcDonald’s, Coca‑Cola, Procter & Gamble, Nike, Walmart
HealthcareJohnson & Johnson, Merck, Amgen
EnergyChevron
CommunicationsWalt Disney
MaterialsDow Inc.

IBM’s ‘block of flats’ chip design pushes Moore’s Law into new territory

IBM chip stack design

IBM’s latest research breakthrough – a sub‑1nm chip architecture built like a “block of flats” – marks one of the most ambitious attempts yet to stretch Moore’s Law beyond its natural limits.

The company claims its new NanoStack design can pack almost 100 billion transistors onto a fingernail‑sized chip, a density that would have been unthinkable even a decade ago.

In early tests, the prototype delivered 50% higher performance and 70% better energy efficiency than IBM’s own 2nm technology, signalling a potential generational leap in computing power.

Moore’s Law at 50 years

For more than half a century, Moore’s Law – the observation that transistor counts double roughly every two years – has shaped the trajectory of the semiconductor industry.

But as transistors approach atomic scales, the physics has become unforgiving. Leakage, heat, and quantum effects increasingly threaten the neat exponential curve that once defined progress.

The industry’s response has been to move vertically: instead of squeezing more transistors across a flat surface, designers are now building upwards.

Verical stacking

IBM’s NanoStack takes this vertical shift to an extreme. Rather than simply elongating transistor structures, the company has begun stacking entire sheets of transistors on top of one another, creating a skyscraper‑like arrangement.

Professor Alan Woodward of the University of Surrey reportedly likens the shift to replacing a city of houses with a 100‑storey tower block – a vivid contrast to the 30–50‑storey equivalents being pursued by rivals such as Samsung and Intel.

The approach is bold, but it comes with engineering hazards. Heat rises through the stack, threatening performance and reliability. Layers that are too thin risk transistors failing to switch off cleanly, undermining the chip’s logic.

Obstacles

These are not trivial obstacles, and IBM acknowledges that commercial production remains several years away.

Yet the company argues that the architectural shift is essential if computing is to keep pace with the demands of AI, cloud workloads, and energy‑constrained data centres.

If NanoStack proves manufacturable at scale, it could represent the most significant extension of Moore’s Law since the industry moved from planar to FinFET designs.

The broader question is whether this vertical strategy can deliver multiple generations of improvement, or whether it is the final flourish before the industry must abandon transistor‑count metrics altogether.

For now, IBM has injected fresh momentum into a field long assumed to be running out of road – and reminded the industry that Moore’s Law may bend, but it is not yet broken.

Moore’s Law states

Moore’s Law is the principle that the number of transistors on a microchip doubles roughly every two years, leading to continual increases in computing power and efficiency.

Quantum Computing’s Breakthrough Moment Puts Data Centres under the Spotlight

Quantum Computing Advances

A quiet but consequential shift is taking place across the global technology landscape: quantum computing is no longer a distant scientific ambition but an emerging commercial reality.

A new wave of breakthroughs is accelerating timelines, and data‑centre operators — already strained by the explosive growth of AI workloads — are being forced to rethink their infrastructure from the ground up.

The latest reporting highlights how this ‘quantum moment’ is reshaping priorities across the sector.

Advancements in Quantum computing

For years, quantum computing has been framed as a long‑term bet, with practical applications perpetually a decade away. That narrative is now being challenged.

Advances in qubit stability, error‑correction techniques and *photonic architectures are pushing the field closer to machines capable of solving commercially meaningful problems.

Industry leaders increasingly argue that hybrid quantum–classical systems will begin appearing inside data centres before the end of the decade, creating a new class of high‑value workloads.

This shift is happening at a time when data centres are already under unprecedented strain. The rapid adoption of generative AI has driven demand for power, cooling and specialised silicon to levels few operators anticipated.

Layered complexity

Quantum computing adds a new layer of complexity: these machines require ultra‑stable environments, extreme cooling and highly specialised networking.

As a result, data‑centre design is entering a new phase, with operators exploring everything from cryogenic‑ready layouts to quantum‑secure communication links.

The strategic implications are significant. Hyperscalers are positioning themselves early, investing in quantum‑safe encryption, photonic interconnects and experimental quantum modules that can be slotted into existing facilities.

Objective

The goal is to ensure that when quantum hardware becomes commercially viable, the supporting infrastructure is already in place.

This mirrors the early days of cloud computing, when capacity was built ahead of demand — a gamble that ultimately paid off.

Yet uncertainty remains. Some analysts caution that full‑scale commercialisation could still be decades away, pointing to slow revenue growth and persistent engineering challenges.

Even so, the direction of travel is clear: quantum computing is moving out of the lab and into the strategic planning of the world’s largest data‑centre operators.

If AI defined the last wave of infrastructure investment, quantum may define the next. And for an industry already racing to keep up, the clock has started ticking.

Explainer

What are Photonic Architectures?

Photonic architectures in quantum computing refer to systems that use light particles (photons) as the fundamental units of quantum information — instead of electrons or superconducting circuits.

These architectures are gaining traction because photons offer several unique advantages:

Key Features of Photonic Quantum Architectures

FeatureDescription
Qubits via photonsQuantum bits are encoded in properties of light, such as polarisation or phase.
Room-temperature operationUnlike superconducting systems, photonic setups often don’t require cryogenic cooling.
Low noise and decoherencePhotons are less prone to environmental interference, improving stability.
Modularity and scalabilityPhotonic systems can be built using modular optical components, ideal for scaling.

The New Wave of AI Anxiety: Why Every Sector Suddenly Feels Exposed

AI related job adjustment

A curious shift has taken place over the past year. The fear of AI ‘taking over’ is no longer confined to software engineers, coders, or the legal and financial professions.

It has spilled into transport logistics, estate agency, recruitment, customer service, and even the once‑untouchable world of creative work.

Anxiety spreads

The anxiety is spreading horizontally across the economy rather than vertically within a single industry — and that tells us something important about where we are in the technological cycle.

At the heart of this unease is a simple realisation: AI is no longer a specialised tool. It is becoming a general‑purpose capability, much like electricity or the internet.

When a technology can be applied to almost any workflow, the boundaries between ‘safe’ and ‘at risk’ jobs dissolve.

Estate agents see AI systems that can generate listings, negotiate pricing models, and automate client follow‑ups. Logistics managers watch algorithms optimise routes, staffing, and inventory with a precision no human team can match.

Even white‑collar professionals, once insulated by complexity and regulation, now face AI systems capable of drafting contracts, analysing case law, or producing financial models in seconds.

This broadening of impact is what’s fuelling the current wave of concern. It’s not that AI is replacing everyone — it’s that it could plausibly reshape the value chain in every sector.

Axis shift

For the stock market, this shift has created a two‑speed economy. Companies building AI infrastructure — chips, cloud platforms, foundation models — are being rewarded with valuations that assume long‑term dominance.

Meanwhile, firms whose business models rely on labour‑intensive processes are being quietly repriced. Investors are asking a new question: Which companies can integrate AI fast enough to defend their margins? Those that can’t risk being treated like legacy utilities.

But the story isn’t simply about winners and losers. The diffusion of AI across industries also creates a multiplier effect.

Productivity gains in logistics lower costs for retailers; smarter estate agency tools accelerate housing transactions; automated legal drafting reduces friction for start‑ups. Each improvement compounds the next.

AI taking over?

The fear, then, is partly a misunderstanding. AI isn’t ‘taking over’ — it’s infiltrating. It is dissolving inefficiencies, redrawing job descriptions, and forcing companies to rethink what they actually do.

The stock market has already priced in the first wave of this transformation. The second wave — where every sector becomes an AI‑enabled sector — is only just beginning.

Are U.S. Markets in an ‘Everything Bubble’?

U.S. Stock Everything Bubble?

The phrase ‘everything bubble‘ has gained traction among investors and commentators who fear that multiple asset classes in the United States are simultaneously overvalued.

Unlike past episodes where excess was concentrated in one sector—such as technology in the late 1990s or housing in the mid‑2000s—the current concern is that equities, property, and credit markets are all inflated together, leaving little room for error.

Equities are the most visible part of the story. Major U.S. indices have surged to record highs, driven by enthusiasm for artificial intelligence, cloud computing, and digital infrastructure.

Valuations in leading technology firms are stretched, with price‑to‑earnings ratios far above historical averages. Critics argue that investors are extrapolating future growth too aggressively, while ignoring the risks of higher interest rates and slowing global demand.

Market breadth has also narrowed, with a handful of companies accounting for most of the gains, a pattern often seen before corrections.

Housing

Housing provides another layer of concern. Despite higher mortgage rates, U.S. home prices remain elevated, supported by limited supply and strong demand in metropolitan areas.

This resilience has surprised analysts, but it also raises the question of sustainability. If borrowing costs remain high, affordability pressures could eventually weigh on the market, exposing households to financial stress.

Credit markets

Credit markets add a third dimension. Corporate debt issuance has slowed, and investors have become more selective, demanding higher yields to compensate for risk. Some deals have been pulled altogether, signalling caution beneath the surface.

When credit tightens, it often foreshadows broader economic weakness, as companies struggle to refinance or fund expansion.

Yet it would be simplistic to declare that everything is a bubble. The rapid adoption of AI and accelerated computing reflects genuine structural change, not mere speculation.

Demand for advanced chips and data centres is tangible, and some firms are generating real cash flows from these trends. Similarly, housing shortages are rooted in years of under‑building, suggesting that supply constraints, rather than speculative mania, are keeping prices high.

The truth may lie in between. U.S. markets are undeniably expensive, and vulnerabilities are widespread.

But not all sectors are equally fragile, and some are underpinned by lasting shifts in technology and demographics.

Investors should therefore resist blanket labels and instead distinguish between genuine transformation and speculative excess.

In doing so, they can navigate a landscape that is frothy in places, but not uniformly illusory.

Nvidia Q3 results were very strong – but does the AI bubble reside elsewhere – such as with the debt driven AI data centre roll out – and crossover company deals?

AI debt

Nvidia’s Q3 results show strength, but the real risk of an AI bubble may lie in the debt-fuelled data centre boom and the circular crossover deals between tech giants.

Nvidia’s latest quarterly earnings were nothing short of spectacular. Revenue surged to $57 billion, up 62% year-on-year, with net income climbing to nearly $32 billion. The company’s data centre division alone contributed $51.2 billion, underscoring how central AI infrastructure has become to its growth.

These figures have reassured investors that Nvidia itself is not the weak link in the AI story. Yet, the question remains: if not Nvidia, where might the bubble be forming?

Data centre roll-out

The answer may lie in the debt-driven expansion of AI data centres. Building hyperscale facilities requires enormous capital outlays, not only for GPUs but also for power, cooling, and connectivity.

Many operators are financing this expansion through debt, betting that demand for AI services will continue to accelerate. While Nvidia’s chips are sold out and cloud providers are racing to secure supply, the sustainability of this debt-fuelled growth is less certain.

If AI adoption slows or monetisation lags, these projects could become overextended, leaving balance sheets strained.

Crossover deals

Another area of concern is the crossover deals between major technology companies. Nvidia’s Q3 was buoyed by agreements with Intel, OpenAI, Google Cloud, Microsoft, Meta, Oracle, and xAI.

These arrangements exemplify a circular investment pattern: companies simultaneously act as customers, suppliers, and investors in each other’s AI ventures.

While such deals create momentum and headline growth, they risk masking the true underlying demand.

If much of the revenue is generated by companies trading capacity and investment back and forth, the market could be inflating itself rather than reflecting genuine end-user adoption.

Bubble or not to bubble?

This dynamic is reminiscent of past bubbles, where infrastructure spending raced ahead of proven returns. The dot-com era saw fibre optic networks built faster than internet businesses could monetise them.

Today, AI data centres may be expanding faster than practical applications can justify. Nvidia’s results prove that demand for compute is real and immediate, but the broader ecosystem may be vulnerable if debt levels rise and crossover deals obscure the true picture of profitability.

In short, Nvidia’s strength does not eliminate bubble risk—it merely shifts the spotlight elsewhere. Investors and policymakers should scrutinise the sustainability of AI infrastructure financing and the circular nature of tech partnerships.

The AI revolution is undoubtedly transformative, but its foundations must rest on genuine demand rather than speculative debt and self-reinforcing deals.

AWS Outage Reveals Fragility of Global Cloud Dependency

Amazon services go dark

It was just one week ago on Monday 20th October 2025, Amazon Web Services (AWS) experienced a major outage that rippled across the digital world, disrupting operations for millions of users and businesses.

The incident, which originated in AWS’s US-East-1 region, was reportedly traced to DNS resolution failures affecting DynamoDB—one of AWS’s core database services.

This technical fault triggered cascading issues across EC2, network load balancers, and other critical infrastructure, leaving many services offline for hours.

The impact was immediate and widespread. Major consumer platforms such as Snapchat, Reddit, Disney+, Canva, and Ring doorbells went dark.

Financial services including Venmo and Robinhood faltered, while airline customers at United and Delta struggled to access bookings. Even British government portals like Gov.uk and HMRC were affected, underscoring the global reach of AWS’s infrastructure.

World leader

AWS is the world’s leading cloud provider, commanding roughly one-third of the global market—well ahead of Microsoft Azure and Google Cloud.

Millions of companies, from startups to multinational corporations, rely on AWS for everything from data storage and virtual servers to machine learning and content delivery.

Its services underpin critical operations in healthcare, education, retail, logistics, and media. When AWS stumbles, the internet itself feels the tremor.

20 Prominent Companies Affected by the AWS Outage (20th Oct 2025)

SectorCompany NameImpact Summary
E-commerceAmazonInternal systems and Seller Central offline
Social MediaSnapchatApp outages and delays
StreamingDisney+Service interruptions
NewsRedditPartial outages, scaling issues
Design ToolsCanvaHigh error rates, reduced functionality
Smart HomeRingDevice connectivity issues
FinanceVenmoTransaction delays
FinanceRobinhoodTrading disruptions
AirlinesUnited AirlinesBooking and check-in issues
AirlinesDelta AirlinesReservation access problems
TelecomT-MobileIndirect service disruptions
GovernmentGov.ukPortal access issues
GovernmentHMRCService delays
BankingLloyds BankOnline banking affected
ProductivityZoomMeeting access issues
ProductivitySlackMessaging delays
EducationCanvasAssignment submissions disrupted
CryptoCoinbaseUser access failures
GamingRobloxServer outages
GamingFortniteGameplay interruptions

This outage wasn’t the result of a cyberattack, but rather a technical fault in one of Amazon’s main data centres. Yet the consequences were no less severe.

Amazon’s own operations were disrupted, with warehouse workers unable to access internal systems and third-party sellers locked out of Seller Central.

Canva reported ‘significantly increased error rates’. while Coinbase and Roblox cited cloud-related failures.

The incident serves as a stark reminder of the risks inherent in centralised cloud infrastructure. As digital life becomes increasingly dependent on a handful of providers, the potential for systemic disruption grows.

A single point of failure can cascade across industries, affecting everything from classroom assignments to emergency services.

AWS has since restored normal operations and promised a detailed post-event summary. But for many, the outage has reignited questions about resilience, redundancy, and the wisdom of placing so much trust in a single cloud giant.

In the age of digital interdependence, even a brief lapse can feel like a global blackout.

Oracle Cloud reportedly to deploy 50,000 AMD AI chips, signalling direct competition with Nvidia

Oracle Cloud AI

Oracle Bets Big on AMD AI Chips, Challenging Nvidia’s Dominance

Oracle Cloud Infrastructure has announced plans to deploy 50,000 AMD Instinct MI450 graphics processors starting in the second half of 2026, marking a bold strategic shift in the AI hardware landscape.

The move signals a direct challenge to Nvidia’s long-standing dominance in the data centre GPU market, where it currently commands over 90% market share.

AMD’s MI450 chips, unveiled earlier this year, are designed for high-performance AI workloads and can be assembled into rack-sized systems that allow 72 chips to function as a unified engine.

This architecture is tailored for inferencing tasks—an area Oracle believes AMD will excel in. ‘We feel like customers are going to take up AMD very, very well’, reportedly said Karan Batta, Oracle Cloud’s senior vice president.

The announcement comes amid a broader realignment in the AI ecosystem. OpenAI, historically reliant on Nvidia hardware, has recently inked a multi-year deal with AMD involving processors requiring up to 6 gigawatts of power.

If successful, OpenAI could acquire up to 10% of AMD’s shares, further cementing the chipmaker’s role in next-generation AI infrastructure.

Oracle’s pivot also reflects its ambition to compete with cloud giants like Microsoft, Amazon, and Google. With a reported five-year cloud deal with OpenAI potentially worth $300 billion, Oracle is positioning itself not just as a capacity provider but as a strategic AI enabler.

While Nvidia remains a formidable force, Oracle’s investment in AMD chips underscores a growing appetite for alternatives.

As AI demands scale, diversity in chip supply could become a competitive advantage—especially for enterprises seeking flexibility, cost efficiency, and innovation beyond the Nvidia ecosystem.

The AI arms race is far from over, but Oracle’s latest move suggests it’s no longer content to play catch-up. It’s aiming to redefine the rules.

With all the new AI tech arriving in the new AI data centres – what is happening to the old tech it is presumably replacing?

AI - dirty little secret or clean?

🧠 What’s Happening to the Old Tech?

Shadow in the cloud

🔄 Repurposing and Retrofitting

  • Many traditional CPU-centric server farms are being retrofitted to support GPU-heavy or heterogeneous architectures.
  • Some legacy racks are adapted for edge computing, non-AI workloads, or low-latency services that don’t require massive AI computing power.

🧹 Decommissioning and Disposal

  • Obsolete hardware—especially older CPUs and low-density racks—is being decommissioned.
  • Disposal is a growing concern: e-waste regulations are tightening, and sustainability targets mean companies must recycle or repurpose responsibly.

🏭 Secondary Markets and Resale

  • Some older servers are sold into secondary markets—used by smaller firms, educational institutions, or regions with less AI demand.
  • There’s also a niche for refurbished hardware, especially in countries where AI infrastructure is still nascent.

🧊 Cold Storage and Archival Use

  • Legacy systems are sometimes shifted to cold storage roles—archiving data that doesn’t require real-time access.
  • These setups are less power-intensive and can extend the life of older tech without compromising performance.

⚠️ Obsolescence Risk

  • The pace of AI innovation is so fast that even new data centres risk early obsolescence if they’re not designed with future workloads in mind.
  • Rack densities are climbing—from 36kW to 80kW+—and cooling systems are shifting from air to liquid, meaning older infrastructure simply can’t keep up.

🧭 A Symbolic Shift

This isn’t just about servers—it’s about sovereignty, sustainability, and the philosophy of obsolescence. The old tech isn’t just being replaced; it’s being relegated, repurposed, or ritually retired.

There’s a tech history lesson unfolding about digital mortality, and how each new AI cluster buries a generation of silicon ancestors.

Infographic: ‘New’ AI tech replacing ‘Old’ tech in data centres

🌍 The Green Cost of the AI Boom

Energy Consumption

  • AI data centres are power-hungry beasts. In 2023, they consumed around 2% of global electricity—a figure expected to rise by 80% by 2026.
  • Nvidia’s H100 GPUs, widely used for AI workloads, draw 700 watts each. With millions deployed, the cumulative demand is staggering.

💧 Water Usage

  • Cooling these high-density clusters often requires millions of litres of water annually. In drought-prone regions, this is sparking local backlash.

🧱 Material Extraction

  • AI infrastructure depends on critical minerals—lithium, cobalt, rare earths—often mined in ecologically fragile zones.
  • These supply chains are tied to geopolitical tensions and labour exploitation, especially in the Global South.

🗑️ E-Waste and Obsolescence

  • As new AI chips replace older hardware, legacy servers are decommissioned—but not always responsibly.
  • Without strict recycling protocols, this leads to mountains of e-waste, much of which ends up in landfills or exported to countries with lax regulations.

The Cloud Has a Shadow

This isn’t just about silicon—it’s about digital colonialism, resource extraction, and the invisible costs of intelligence. AI may promise smarter sustainability, but its infrastructure is anything but green unless radically reimagined.

⚡ The Energy Cost of Intelligence

🔋 Surging Power Demand

  • AI data centres are projected to drive a 165% increase in global electricity consumption by 2030, compared to 2023 levels.
  • In the U.S. alone, data centres could account for 11–12% of total power demand by 2030—up from 3–4% today.
  • A single hyperscale facility can draw 100 megawatts or more, equivalent to powering 350,000–400,000 electric vehicles annually.
AI and Energy supply

🧠 Why AI Is So Power-Hungry

  • Training large models like OpenAI Chat GPT or DeepSeek requires massive parallel processing, often using thousands of GPUs.
  • Each AI query can consume 10× the energy of a Google search, according to the International Energy Agency.
  • Power density is rising—from 162 kW per square foot today to 176 kW by 2027, meaning more heat, more cooling, and more infrastructure.

🌍 Environmental Fallout

  • Cooling systems often rely on millions of litres of water annually. For example, in Wisconsin, two AI data centres will consume 3.9 gigawatts of power, more than the state’s nuclear plant.
  • Without renewable energy sources, this surge risks locking regions into fossil fuel dependency, raising emissions and household energy costs. We are not ready for this massive increase in AI energy production.

Just how clean is green?

The Intelligence Tax

This isn’t just about tech—it’s about who pays for progress. AI promises smarter cities, medicine, and governance, but its infrastructure demands a hidden tax: on grids, ecosystems, and communities.

AI is a hungry beast, and it needs feeding. The genie is out of the bottle!

How frothy is the AI data centre market for investors?

AI market froth?

Nvidia investors have been on a rocket ride to the stars. But recently they have come back down to Earth, and it has become more of a roller coaster ride.

Benefiting significantly from the artificial intelligence surge, Nvidia’s market cap has increased approximately ninefold since late 2022 – a massive market cap gain.

However, after achieving a peak in June 2024 and momentarily claiming the title of the world’s most valuable public company, Nvidia then experienced close to a 30% decline in value over the subsequent seven weeks, resulting in an approximate $800 billion loss in market capitalisation.

Currently, the stock is experiencing a rally, bringing it within approximately 6% of its all-time peak. The chipmaker surpassed the $3 trillion market cap milestone in early June 2024, aligning with Microsoft and Apple. The question remains whether the company can reclaim and sustain that title.

Investors are closely monitoring Nvidia’s forecast for the October quarter, with the company anticipated to report a growth of approximately 75%. Positive guidance would imply that Nvidia’s affluent clients continue to invest heavily in AI development, whereas a lacklustre forecast might suggest that infrastructure investment is becoming excessive.

Should there be any signs of diminishing demand for AI or if a major cloud customer is reducing spending, it could lead to a notable decline in revenue.

Microsoft shares drop on cloud miss as Azure revenue disappoints

In the cloud

Microsoft reported better-than-expected earnings and revenue for Q4

In extended trading on 30th July 2024, the stock experienced a quick decline as attention was drawn to the less-than-expected Azure revenue, despite management’s forecast for growth in the upcoming quarters.

The company’s total revenue saw a 15% increase compared to the previous year.

Despite surpassing earnings and revenue expectations, Microsoft’s shares dropped by up to 7% in extended trading on Tuesday, with investors concentrating on the underwhelming cloud revenue. However, executives offered a positive outlook, anticipating an acceleration in cloud growth during the first half of 2025.

Microsoft one day chart 30th July 2024

Microsoft one day chart 30th July 2024

Microsoft’s cloud division holds significant interest for investors, as it competes with Amazon Web Services (AWS) and Google in the artificial intelligence (AI) work arena. These three tech giants are pouring substantial resources into enhancing AI capabilities, aiming to attract both startups and established companies as generative AI technology swiftly progresses.

For Amazon, AWS has served as a vital profit centre for the past ten years.

Wiz dumps $23 billion deal with Google -reportedly to pursue IPO

Online security

Wiz has apparently walked away from a deal with Google that would have valued the company at $23 billion.

The deal would have nearly doubled the $12 billion valuation of the startup from its most recent round of funding.

CEO of WIZ Assaf Rappaport told employees the company would pursue an IPO as originally planned.

Wiz was founded in 2020 and has grown rapidly. The company had been targeting an IPO as recently as May 2024. The business hit $100 million in annual recurring revenue after 18 months and reached $350 million last year.

Wiz’s cloud security products offer prevention, active detection and response, a portfolio that’s appealed to large firms and would have helped Google compete with Microsoft, which also sells security software.

One to watch for a potential future IPO.

Energy hungry data centre power solution

AI data centre

The use of nuclear reactors for data centres is a controversial and complex topic that has both advantages and disadvantages

Nuclear reactors can provide a reliable, stable, and carbon-free source of electricity for power-hungry data centres, which are essential for the operation of various applications, such as artificial intelligence (AI).

Grid overload

Nuclear reactors can also reduce the dependence on the existing grid, which may be vulnerable to blackouts, fluctuations, or cyberattacks. On the other hand, nuclear reactors require a high initial investment, as well as strict safety and regulatory standards. Nuclear reactors also pose potential risks of radiation, waste disposal, and proliferation. Moreover, nuclear reactors may not be suitable for all locations, as they may face public opposition, environmental concerns, or geopolitical issues.

Small Modular Reactor (SMR)

One of the possible solutions to these challenges is to use small modular reactors (SMRs), which are advanced reactors with about a third of the power generation of a traditional, large nuclear plant. SMRs are designed to be more flexible, scalable, and cost-effective than conventional reactors, as they can be built off-site and transported to the desired location. SMRs can also be integrated with renewable energy sources, such as solar or wind, to create a hybrid system that can balance the power demand and supply.

However, the technology of SMRs is still in its early stages of development and deployment, and there are currently no data centres in the world that use built-in nuclear reactors. Therefore, it remains to be seen whether nuclear reactors will become a common or viable option for future data centres. The decision to use nuclear reactors for data centres should be based on a careful evaluation of the benefits and risks, as well as the alternatives and trade-offs, of each specific case.

It has been calculated that a ‘norma’ data centre (whatever that is), needs 32 megawatts of power flowing into the building. For an AI data centre, it’s closer to 80 megawatts.

AI systems are using all this extra electricity simply because they are doing so much more processing than standard computing. They are chewing through far more data.

As AI continues to develop, so too will the power requirement needed to run these monsters.

Microsoft closes at all-time high

Microsoft closes at all-time high

Microsoft ended Tuesday’s trading session at a record high of $360.53, following fresh optimism about growth from a key partner in artificial intelligence (AI). The increase gives the company a market value of about $2.68 trillion.

At a tech event on Monday 6th November 2023, Microsoft’s AI partner, OpenAI, announced a batch of updates, including price cuts and plans to allow people to make custom versions of the ChatGPT chatbot.

Microsoft CEO Satya Nadella attended and emphasized that developers building applications with OpenAI’s tools could get to market quickly by deploying their software on Microsoft’s Azure cloud infrastructure.

Microsoft has invested a reported $13 billion in OpenAI, which has granted Microsoft an exclusive licence on OpenAI’s GPT-4 large language model that can generate human-like prose in response to a few words of text.

Chatbot
Fictitious AI robot learning from a digital human online

Last week, Microsoft announced the release of an AI add-on for its Office productivity app subscriptions and an assistant in Windows 11, both of which rely on OpenAI models.

The future is looking bright for Microsoft right now.

IBM pivots to AI – STOCK WATCH

IBM

An old well established and trusted tech brand pivoting to AI that has a high dividend yield is IBM, which has been around for more than a century and is known for both its hardware and software products. 

IBM is investing heavily in AI, cloud computing, and quantum computing, and has recently acquired several AI start-ups, such as Instana, Turbonomic, and Waeg. 

IBM also has a partnership with OpenAI, one of the leading AI research organizations, to provide cloud infrastructure for its AI models.

Investors who love IBM expect the company to grow its earnings by around 10% annually over the next five years. Investors were also impressed with IBM’s dividend yield, which is currently around 4.5%. Dividends are a great way to generate passive income.

IBM is not the only tech company that is pivoting to AI. Google, Microsoft, and Anthropic are competing in the field of generative AI, which can create text, images, music, and more from natural language prompts. 

Integrate generative AI

These companies are attempting to integrate generative AI into their products and services, such as search engines, maps, word processors, office applications, chatbots, and more. Generative AI is seen as a game-changer for many industries and applications, and could potentially disrupt the dominance of Big Tech.

Legacy companies can pivot to a platform model, which is a business strategy that connects producers and consumers of value through a digital interface. Platform companies like Facebook, Amazon, Google, and Tencent have created value at stunning rates, and have grown rapidly and own large market shares. 

IBM mainframe from the 1970’s

Legacy companies can leverage their existing systems, such as customer relationships, data, and brand recognition, to create platforms that offer impressive and immersive products and services. 

Other successful platform pivots are Disney+, which transformed Disney from a media producer to a media platform; Nike+, which connected Nike’s physical products with digital services; and John Deere, which created a platform for precision agriculture.

Cybersecurity

Hack attack!

Cybersecurity is a very important and relevant topic in today’s world. It refers to the practice of protecting systems, networks, and programs from digital attacks that can harm individuals and organizations.

Cyberattacks will all have malicious intent, such as accessing, changing, or destroying sensitive information; extorting money from users via ransomware; or interrupting normal business processes.

Cybersecurity aims to prevent or mitigate these attacks by using various technologies, measures, and practices.

There are many types of cybersecurity, depending on the domain or layer of IT infrastructure that needs to be protected.

Critical infrastructure security

This protects the computer systems, applications, networks, data and digital assets that a society depends on for national security, economic health and public safety. For example, the power grid, the water supply, the transportation system, the health care system, etc. 

In the United States, there are some guidelines and frameworks for IT providers in this area, such as the NIST cybersecurity framework and the CISA guidance.

Network security

This prevents unauthorized access to network resources and detects and stops cyberattacks and network security breaches in progress. For example, firewalls, antivirus software, encryption, VPNs, etc. Network security also ensures that authorized users have secure access to the network resources they need, when they need them.

Application security

This protects applications from cyberattacks by ensuring that they are designed, developed, tested, and maintained with security in mind. For example, code reviews, vulnerability scanning, penetration testing, secure coding practices, etc. Application security also involves educating users about safe and responsible use of applications.

Cyberattacks will all have malicious intent, such as accessing, changing, or destroying sensitive information; extorting money from users via ransomware; or interrupting normal business processes.

There are many more types of cybersecurity, such as cloud security, endpoint security, data security, identity and access management (IAM), etc. Each type of cybersecurity has its own challenges and solutions.

Companies to watch

Cybersecurity companies such as CrowdStrike, Okta, Zscaler and Palo Alto Networks are valuable assets with businesses willing to pay good money to protect against hackers.

Zscaler

Palo Alto Networks

Crowdstrike

Okta

NOTE: Always do your own very careful research – none of these ‘suggestions’ are ‘recommendations’.

Remember: RESEARCH! RESEARCH! RESEARCH!

Nvidia’s stock at record high after Google AI deal

AI microchip

Nvidia shares rose 4.2% Tuesday 29th August 2023 to close at a record high, after the company announced a partnership with Google that could expand distribution of its artificial intelligence technology (AI).

The stock’s bountiful run continued, now up 234% in 2023, making it by far the best performer in the S&P 500. Facebook parent Meta is second in the index, up 148% so far this year.

The record close comes less than a week after the company said quarterly revenue doubled from a year earlier and gave a forecast indicating that sales this period could rise 170% on an annual basis. The day after the better-than-expected earnings report, the stock climbed to a record intraday high of $502.66 before declining later in the afternoon.

Nvidia’s business is booming because its graphics processing (GPU’s) are being gobbled up by cloud companies, government agencies and startups to train and deploy generative AI models like the technology deployed in OpenAI’s ChatGPT as fasta as Nvidia can make them.

NVIDIA stock chart

Nvidia announcment

On Tuesday 29th August 2023, Nvidia CEO Jensen Huang appeared at a Google conference to announce an AI agreement between the two companies.

Through the partnership, Google’s cloud customers will have greater access to technology powered by Nvidia’s powerful H100 GPUs.

‘Our expanded collaboration with Google Cloud will help developers accelerate their work with infrastructure, software and services that supercharge energy efficiency and reduce costs’, the Nvidia CEO reportedly said in a blog post.

Nvidia’s GPUs are also available on competing cloud platforms from Amazon and Microsoft.