The AI bubble burst is probably not a question of if – it’s more a forecast of when, how hard the fall will be, how long the recovery will take, and what the business impacts will be during and after the shakeout.

To be fair, some say the bubble is confined to a few companies, while others argue the AI bubble may deflate, but not bust.

I survived the bursting internet bubble in 2001 as a startup CTO, and then the financial meltdown in 2008 as a transformational CIO. CIOs should read the tea leaves and consider the signs, then develop a plan to manage potential upcoming turbulence across markets, sentiment, supply chains, and financial impacts.

CIOs should reflect on their AI progress

Many CIOs are delivering business value with AI agents, and top companies are shaping the future of work with AI capabilities. They’re developing their organization’s AI brains, investing in knowledge management, scaling rollouts with AI agent orchestration platforms, and engaging the workforce with vibe no-code capabilities. 

Other CIOs are investing in AI, but delivering value is a work in progress. Some give in to AI’s hype by driving chaotic experimentation or underinvesting in AI governance. Many will make mistakes when deploying AI agents, creating new forms of AI debt and AI cost debt.

Wherever CIOs are on the AI adoption spectrum, an AI bubble burst or deflation, will impact strategy, priorities, investment levels, and employee sentiment.

The evidence: 10 signs of an AI bubble near bursting

I partnered with Claude on the research; I hypothesized about the signs, and it found supporting evidence. It challenged me on two of them, which I researched and restated. If you want to see Claude’s unedited report, reach out to me and I’ll share it with you.

Here’s a summary of the evidence. You decide what’s going to be the pin.

1. Investors are losing patience with inflated AI stock prices. CNBC reports AI is driving a looming market correction, while Fortune says the AI math doesn’t make sense.
2. Enterprises respond to AI’s rising costs and ROI gaps. The ROI pressure is surging, and recent research continues to show the gap: Infosys reports that 72% have scaled fewer than one-fourth of their AI pilots, KPMG reports that only 7% of leaders report having established ROI from AI, and McKinsey says record AI spending can’t move EBIT for 94% of enterprises.
3. VC bets are concentrated in big AI. In H1 2026, rounds of $100m+ took 87.5% of US venture dollars, mostly in AI. In the EU, mega-rounds accounted for more than half of deal value in Q2, with €26.5 billion in the AI sector.
4. AI’s financing boom just lost its guardrails. CNBC reports that AI’s infrastructure boom is becoming more leveraged and harder to track, and Fortune says the AI debt orgy can’t last forever, while hidden borrowing has exploded to $1.65 trillion. Michael Burry took notice, saying a big threat to the AI boom is lurking in private credit, while Business Standard reports that the AI borrowing binge is helping drive US government bond yields.
5. Growing public backlash against data centers. The New York Times reports on how voters are responding, while The Washington Post labels data centers as toxic politics.
6. Growing resentment as AI impacts the workforce. I wrote about how 2026 college grads are pissed off about AI, while Stanford reported that the AI employment gap for young workers is 19%. CNBC reported that employers cited AI as the primary reason for almost 40% of May’s announced job cuts, up from 7% in January.
7. Chip scarcity shifts from AI to memory. There are multiple supply chain risks,  industry lobbyists are out to claim their share, and the Trump administration is considering a second round of semiconductor tariffs.
8. AI’s energy impacts are now a call to action. Virginia regulators have ordered their top energy utility to charge data centers the direct costs of their electric service, while opening the door for them to pay more of the grid’s transmission costs. PJM wants to cut power to data centers during emergencies before impacting consumers.
9. Major AI security incidents impact more than reputations. While OpenAI’s hack of Hugging Face made the headlines, IBM reported that one-fourth of malicious breaches are AI-enabled and costing $6M on average.
10. The Chinese AI models are a wildcard. Alibaba’s AI model reached 3 billion downloads, and Chinese models are gaining ground on OpenAI and Anthropic, while NVIDIA warns of business risks if the Trump Administration imposes restrictions on AI models.

Four financial triggers (1-4), two on public sentiment (5-6), two around supply and energy impacts (7-8), and two wildcards (9-10) may be the pin.

How CIOs should prep for the ROI reckoning

CIOs don’t have to be futurists or plan for every scenario, but today’s plans must be reviewed against the likelihood of an AI bubble burst.

Here are my seven recommendations for CIOs if there’s a burst or deflation:

1. Double down on digital transformation force multipliers while pruning low-impact initiatives or long-running experiments. Rewrite vision statements for active initiatives reflecting the new realities.
2. Develop a change management plan to address detractors who see the burst as an opportunity to slow down AI initiatives.
3. Catch up on AI governance and data governance, because most organizations have let the AI gold rush drive priorities while guardrails were still being developed.
4. Prioritize addressing AI debt, AI cost debt, and data management debt by developing self-organizing standards and ensuring agile teams align with them.
5. Revisit your AI strategy and develop a communication plan so employees understand that (i) AI is still critical to your business, and (ii) how objectives and priorities have changed.
6. MMBs and SMBs have the opportunity to accelerate their AI initiatives during a shakeup, but must revisit their technology selections for the AI era. Many smaller and industry-specific technology solution providers may not survive an extended downturn.
7. Larger enterprises should revisit their talent strategies, update career planning programs for the jobs most impacted by AI, and double down on AI literacy programs.

I’m not predicting a burst is imminent, just being a pragmatic planner.

---

StarCIO Workshops