🚨 Global markets are in risk-off mode. AI is at the centre of it

The global market mood has changed sharply this week.

What started as a pullback in technology stocks has now spread across equities, emerging markets and commodities. Investors are moving toward safer assets as concerns grow over whether the enormous spending behind the AI boom can generate returns quickly enough to justify the valuations and financing involved.

The MSCI All Country World Index is heading for its seventh straight day of declines , its longest losing streak since September 2023.

And Asia is taking an even bigger hit.

:chart_with_downwards_trend: Asia is bearing the brunt

Asian markets have been hit particularly hard because several of the region’s biggest markets have become closely tied to the AI investment cycle.

  • :taiwan: Taiwan: down 5.2%
  • :kr: South Korea: down 4.4%
  • :jp: Japan: down 4.3%
  • :hong_kong: Hong Kong: down 1.7%
  • :cn: Shanghai: down 1.6%

MSCI’s Asia Pacific equity index is heading for its worst week since March 2020 .

Emerging markets are also under pressure, with the MSCI emerging-markets index now down about 20% from its June peak , putting it close to bear-market territory.

This isn’t just a technology selloff anymore. It is becoming a broader risk-off move .

:robot: So, what changed around AI?

The biggest question investors are asking is no longer whether AI is transformative.

That part is increasingly accepted.

The question is:

Can companies actually make enough money from AI to justify the enormous amount of money being spent to build it?

That concern has become more visible as companies across the AI ecosystem commit hundreds of billions of dollars to computing capacity, chips and data centres.

Anthropic’s IPO was supposed to provide a major confidence boost.

Instead, the reaction has been more complicated.

The company raised $107.3 billion , making it the largest IPO on record and giving it a fully diluted valuation of around $1.5 trillion .

Shares initially jumped strongly, finishing their first trading day 17% above the IPO price .

But by the following session, Anthropic shares fell 4.7% in after-hours trading to $92.60 , slipping below their lowest level during Thursday’s debut.

That tells us something important.

A successful IPO can attract enormous demand for one company while investors remain nervous about the broader AI trade.

:moneybag: The real concern is the cost of the AI boom

Anthropic’s numbers show why investors are paying attention.

The company has seen extraordinary revenue growth, with run-rate revenue reaching about $81 billion as of October 31 .

It also has more than 11,900 customers generating at least $100,000 of annualised revenue.

But there is another number that stands out.

Anthropic’s commitments have climbed to nearly $901 billion , largely linked to computing power and chips.

About $71 billion of those commitments are due in the year through September 2027.

That’s an enormous amount of future spending.

And Anthropic isn’t alone.

Across the technology industry, hyperscalers and AI companies are borrowing and spending aggressively to expand their infrastructure.

So investors are beginning to look beyond revenue growth.

They want to know whether the economics of AI can support the spending required to keep the entire system running.

:bank: Investors are moving toward safety

As confidence in risk assets weakens, money is moving into traditional havens.

The yield on the US 10-year Treasury fell to around 5.02% , while the dollar remained near its strongest level since April 2025.

That is a classic sign of investors becoming more defensive.

At the same time:

  • :coin: Gold fell 2.1%
  • :2nd_place_medal: Silver has fallen around 14% this week
  • :dollar: The dollar is heading for its fourth consecutive weekly gain
  • :chart_with_downwards_trend: Global equities are extending their losses

Interestingly, even commodities that had benefited from the AI infrastructure story are now feeling the pressure.

:coin: Copper is the latest warning sign

Copper had become one of the biggest beneficiaries of the AI infrastructure boom.

The logic was straightforward.

More data centres require more power. More power infrastructure requires more copper.

That helped push copper close to $15,000 a tonne in September .

Now the story is changing.

Copper has fallen roughly 9% this week , trading around $12,213 a tonne and heading toward its worst weekly performance in years.

The decline isn’t only about AI.

Weak Chinese demand, a stronger dollar and changing expectations around US copper tariffs are also weighing on prices.

But the AI connection matters.

Investors had priced in enormous future demand from data centres and infrastructure. Now they are questioning how much of that demand will actually materialise, and how quickly.

:warning: AMD gets caught in the crossfire

AMD is another example of how quickly sentiment can change.

Shares fell 7.7% to $489.19 after Hunterbrook Capital disclosed a short position and raised concerns about AMD’s exposure to China and the risks surrounding its AI business.

AMD rejected the allegations, saying the claims misrepresent its business practices and relationships and that it complies with US export-control requirements.

The bigger issue for investors goes beyond this particular report.

There is growing scrutiny of the relationships between chipmakers, AI companies and their customers.

Investors are increasingly asking whether some of the financing arrangements across the AI ecosystem could create vulnerabilities if AI spending slows.

That makes the market much more sensitive to bad news.

:bar_chart: Nvidia remains a key part of the story

Nvidia is still at the centre of the AI investment cycle.

The company has benefited enormously from the demand for AI chips, but investors are now looking more closely at what happens if AI infrastructure spending eventually slows.

There are also questions around Nvidia’s role in financing and supporting parts of the broader AI ecosystem.

And with the company facing additional scrutiny over its advanced chips and potential access to China, the pressure around the AI trade is coming from multiple directions.

This is why the current selloff feels bigger than a normal technology correction.

Investors are reassessing the entire AI ecosystem, from chips and data centres to financing and the companies developing the models themselves.

:earth_africa: What should investors watch next?

The next few weeks could be important because markets are now looking for evidence rather than promises.

Some of the key questions are:

  • Can AI companies turn massive revenue growth into sustainable profits?
  • Will hyperscalers continue spending at the current pace?
  • Can the infrastructure buildout generate sufficient returns?
  • Will chip demand remain strong if AI capital spending slows?
  • How much financing will the next phase of AI require?
  • Will investors continue paying premium valuations for AI companies?

The biggest test may not come from another AI model launch or another record valuation.

It could come from the balance sheet.

:mag: The bigger picture

The AI story hasn’t suddenly disappeared.

Companies are still investing heavily. Demand for AI infrastructure remains significant, and businesses are continuing to adopt AI tools.

But markets are no longer willing to simply reward spending because it is labelled AI .

That is the important shift.

For much of 2026, investors were willing to look through enormous capital expenditure because they expected even bigger future revenues.

Now they are asking a much harder question:

Where is the return on all this money going to come from?

That question is spreading from AI stocks to semiconductors, copper, emerging markets and even the broader global equity market.

And that is why this week’s selloff deserves attention.

The AI trade may not be over. But the market is clearly becoming much less willing to take its future success for granted.