Expert Warns of Financial Bubble in AI Sector
The rapid expansion of data center infrastructure may lack the revenue to justify current levels of capital expenditure.
Updated on Oct. 26, 2026 in Data Centers
Live Poll
Is now a good time to increase financial exposure to the current artificial intelligence market?
Former Google executive Mo Gawdat warns that the artificial intelligence sector is in a financial bubble driven by unsustainable capital expenditure. He suggests that current investment levels lack sufficient revenue generation and urges investors to shift focus toward hyper-local AI applications.
Why it matters
The massive capital requirements for AI infrastructure currently lack revenue justification, risking a market correction. This volatility mirrors the speculative over-investment of the dot-com era, potentially threatening the stability of major tech-linked assets.
Hyperscale data centers require a minimum of 5,000 servers and 10,000 square feet of space to operate at scale. The proposed Stargate UAE project aims for 5GW of capacity, a significant jump from the nation's total 358MW capacity recorded in 2025.
The players
Mo Gawdat
A former Google executive and commentator on the financial risks of the artificial intelligence sector.
Microsoft
A major provider of cloud infrastructure and AI models that invested $1.5 billion in Emirati firm G42 in 2024.
G42
An Emirati AI holding company focused on large-scale infrastructure and autonomous government service deployment.
SoftBank
A Japanese multinational conglomerate that is currently seeking $100 billion in funding from Gulf investors.
The details
Hyperscale facilities are massive, interconnected compute environments where companies sustain valuations through complex cross-investment webs involving suppliers, investors, and customers. This circular funding model effectively masks the gap between massive capital spending and actual operational revenue. Gawdat argues that this financial structure makes the current infrastructure build-out vulnerable to a burst similar to the market crash of the early 2000s.
Timeline
Late 1990s: Speculative over-investment in internet startups occurred.
Early 2000s: The dot-com market crash occurred.
April 2024: Microsoft invested $1.5 billion in G42.
2025: The UAE total data center capacity was 358MW.
September 2026: Over $9 billion in tech credit default swaps traded weekly.
The Tech Race
Gawdat compares current AI infrastructure investment patterns to the speculative over-investment that led to the dot-com market crash. This assessment contrasts with aggressive expansion plans like the Stargate UAE project, which seeks to reach 5GW of capacity.
Investors and stakeholders should monitor the volume of tech-linked credit default swaps for signals of cooling market sentiment. If a global economic slowdown occurs, users may pivot to more cost-effective AI systems from Chinese providers as firms seek to lower compute costs.
The takeaway
The AI sector faces a structural risk as massive capital outlays remain decoupled from actual revenue growth. Stakeholders should track the weekly volume of tech-linked credit default swaps as a key indicator of market confidence over the coming 12 to 24 months.
Further reading
Explore deeper analysis on the evolving landscape of Data Centers to understand current infrastructure trends.
Source note: This article includes information reported by AGBI.
Live Poll
Is now a good time to increase financial exposure to the current artificial intelligence market?
