One headline by itself doesn’t mean much.
But when several independent signals start pointing in the same direction, it’s worth paying attention.
Last week, Goldman Sachs’ prime brokerage reported hedge funds sold U.S. information technology stocks at an extreme pace, with positioning reaching roughly a minus 4 Z score. That’s the kind of reading you don’t see very often. It suggests aggressive de risking, not routine profit taking.
Then came the Bank for International Settlements.
Its latest report didn’t just talk about AI. It warned about record public debt, financial fragilities, inflation risks, and uncertainty over whether today’s massive AI investment will generate enough long term returns.
Then layer on another development.
Chinese AI companies continue releasing lower cost models that some founders believe could become attractive alternatives for businesses looking to reduce expenses. Whether they gain widespread adoption or not, pricing pressure is becoming part of the conversation.
JUST IN: 🇨🇳 Ex Meta PM and AI founder Xiaoyin Qu says “American and European enterprises will ditch OpenAI and anthropic and adopt Chinese models.” pic.twitter.com/yiIriM84ed
— Whale Insider (@WhaleInsider) June 29, 2026
Even JPMorgan, while remaining constructive on America’s leadership in AI, pointed to growing risks from market concentration, debt sustainability, and geopolitical tensions.
Notice the pattern.
These aren’t all making the same prediction.
They’re highlighting different parts of the same picture.
Here’s the difference in these bubbles:
—Dotcom left fiber
—Railways left tracks—1929 left nothing and we got a Great Depression
AI leaves nothing but empty buildings. GPUs are worthless if you can’t afford to run them, and they age like milk.
This is tulips. https://t.co/dIRppPBh3e
— Jim Stewartson, Decelerationist 🇨🇦🇺🇦🇺🇸 (@jimstewartson) June 29, 2026
Heavy hedge fund selling.
Record debt.
Extreme AI valuations.
Questions about future returns.
Growing competition.
Late cycle economic signals from consumers and the labor market.
None of those automatically mean an AI crash is coming.
But together they suggest professional investors are starting to ask a different question.
Not “Is AI revolutionary?”
Most people already agree it is.
The real question is whether today’s stock prices already assume years of flawless execution, unlimited capital spending, and no meaningful slowdown.
History shows that revolutionary technologies can reshape the world while investors still lose a lot of money if expectations get too far ahead of reality.
That’s what makes this moment interesting.
The debate is no longer about whether AI matters.
It’s about whether the market has already priced in perfection.