AI bubble holds for now, but Forrester urges businesses to prepare exit plans amid economic risks

Tech Giants' Cash Reserves Delay AI Bubble Burst

The AI bubble isn't poised for an imminent collapse, according to Forrester's latest research, but businesses need to start planning their exit strategies now1

1

. The analyst firm attributes this stability to a crucial factor: approximately 90% of AI infrastructure funding comes from financially robust technology companies using their own cash flow, rather than fragile startups dependent on external capital2

2

. Amazon, Google, Microsoft, and Meta are expected to commit more than $740 billion in capital expenditure during 2026 alone, pushing total global AI infrastructure spending past $1 trillion2

2

.

Brian Hopkins, Forrester's emerging technology vice president, emphasizes that while fears of an immediate generative AI bubble burst are premature, economic risks remain concentrated in specific areas. The primary vulnerabilities lie with a small number of builders facing genuine financial strain and circular financing arrangements between AI labs and their backers1

1

. Additionally, major tech firms have raised $326 billion in new tech debt, and nearly $900 billion has been wiped from major tech stocks, creating underlying market volatility2

2

.

The $1.3 Trillion Value Gap Threatening GenAI Market Sustainability

AI infrastructure spending has dramatically outpaced measurable returns, creating significant GenAI market sustainability concerns. Forrester's analysis reveals that current infrastructure investments only break even if AI raises US GDP by 5% by 20351

1

. Published productivity estimates account for merely 2 percentage points of that required growth, leaving approximately $1.3 trillion in value that still needs to materialize from other sources1

1

2

2

.

Management consultancy Bain & Company reached similar conclusions, estimating the AI industry must generate $6 trillion in annual revenue by 2031 to justify ongoing infrastructure investments1

1

2

2

. Other analyses cited by Forrester place the requirement even higher at $8.5 trillion by 2030 to support a 12% return, against current forecasts of just $3.6 trillion2

2

. This substantial gap highlights why achieving lasting AI project value beyond simple productivity gains remains the biggest unknown factor.

Warning Signs Point to Potential AI-Driven Corrections

Forrester has developed its AI Bubble Model specifically to help technology leaders monitor risks and prepare for potential disruptions1

1

. The firm points to several concerning trends that could trigger AI-driven corrections. Token prices have already collapsed as major providers like Anthropic slashed costs in response to competition from open source models1

1

. This pricing pressure could rapidly reset contracts and margins across the industry2

2

.

The venture funding landscape shows additional stress signals, with $519 billion already deployed and OpenAI raising $11 billion toward a $30 billion fundraising target2

2

. If demand disappoints as new capacity comes online, suppliers could cut investment, merge, change prices, or discontinue products entirely. Customers risk finding themselves rebuilding AI plans around suppliers and services that no longer exist1

1

.

Building Your Exit Plan for AI Investments

Forrester advises businesses to actively track whether their AI projects deliver measurable value while limiting exposure to vulnerable suppliers1

1

. Companies should examine the gap between suppliers' contracted business and revenue actually recognized, along with how vendors fund their build-outs. Heavy reliance on outside investment and customer prepayments, as seen with Oracle and CoreWeave, serves as a red flag signaling potential vendor failure risk1

1

.

Enterprises need to stress-test their AI strategies against scenarios including a key vendor announcing end of support within 12 months, or an AI model supplier defaulting on payments to its compute vendor and reducing service-level agreements for enterprise workloads1

1

. Organizations should identify in advance which AI use cases to drop, what data and workloads require migration, and calculate associated costs. Hopkins cautions that oversupply alone won't determine whether a correction arrives—demand will, and that depends entirely on enterprises finding lasting value beyond simple productivity gains1

1

.

Global Financial Institutions Sound Alarm on Financial Fragility

The International Monetary Fund recently began warning that financial markets face increasing exposure to an AI-driven correction that could trigger widespread economic disruption1

1

. The Bank for International Settlements (BIS) raised similar concerns earlier this year, noting that the current AI industry situation bears hallmarks of historical economic bubbles and presents risks to global economic stability1

1

.

Despite these warnings, Hopkins insists organizations shouldn't pull back from AI investments. Instead, they should continue investing where AI delivers demonstrable value, closely monitor the financial health of companies building AI computing infrastructure, and maintain readiness for market corrections1

1

. The shift from whether generative AI can attract money to whether buyers can make it profitable marks a critical transition point for the industry2

2

.