AI Hype Sparks Tech Sell-Off as Bubble Warnings Grow
Added 2025-08-22 05:01:46 +0000 UTCTech and chip stocks tumbled for a second day this week as concerns mounted over the sustainability of the AI boom. At the same time, billionaire investor Howard Marks has cautioned that U.S. markets may be showing early signs of a bubble.
Tech Stocks Under Pressure
Amazon (AMZN) and Apple (AAPL) fell nearly 2%.
Alphabet (GOOGL) slipped about 1%.
Nvidia (NVDA) closed slightly lower after a sharp drop on Tuesday.
AMD (AMD), Broadcom (AVGO), and Micron (MU) also slid, with Micron down ~4%.
AI-linked CoreWeave (CRWV) has now lost over 20% in just five sessions.
Palantir (PLTR) extended its losing streak, down 1%.
The pullback follows an MIT report showing 95% of companies aren’t seeing returns from AI and OpenAI CEO Sam Altman’s admission that AI may indeed be in a bubble phase.
Bubble Talk Heats Up
Howard Marks, co-founder of Oaktree Capital, echoed these concerns, pointing to high valuations, FOMO-driven behavior, and AI hype as warning signs. While he stops short of declaring a full bubble, Marks says markets feel “frothy” and urges discipline.
His warning is backed by major banks:
J.P. Morgan sees policy and geopolitical risks amplifying volatility.
Bank of America warns that rate cuts and tax breaks could inflate growth stock valuations.
Société Générale notes the S&P 500 is at risk of overvaluation if Fed policy eases too much.
The Risk for Investors
Marks and analysts stress avoiding herd mentality and overpaying for hyped assets. Instead, they recommend:
A value-oriented approach focused on fundamentals.
Diversification beyond AI and mega-cap tech.
Caution against narratives justifying sky-high multiples.
Bulls vs. Bears
Not everyone is cautious. Wedbush’s Dan Ives insists the AI bull cycle is intact for 2–3 more years, citing trillions being spent on AI infrastructure.
But Marks reminds investors that past bubbles—from dot-com to housing—started with “a kernel of truth” that spiraled into excess.
What’s Next
The key test comes Aug. 27, when Nvidia reports earnings. Its results will likely shape whether AI enthusiasm stabilizes or bubble fears deepen.
