CITIC SEC: Three Potential Scenarios for the Short-to-Medium-Term Trajectory of the "AI Bubble"

Given the increasing scale of AI investments and their uncertain returns, discussions about an "AI bubble" have become unavoidable in the market. Based on key factors such as algorithmic advancements, core companies' business strategies, and macroeconomic liquidity, CITIC SEC predicts three possible scenarios for the AI industry over the next 12 months. The baseline scenario (60% probability) involves OpenAI facing operational challenges and a slowdown in AI investment. The other two scenarios—short-term breakthroughs in AI algorithms (20% probability) and a resurgence of U.S. inflation leading to a bubble burst (20% probability)—are considered less likely.

### The AI Bubble: An Unavoidable Market Topic Since October, as market risk appetite has declined, investors have begun rationally assessing long-term AI investment sustainability, returns, application timelines, and potential bubbles. CITIC SEC has previously analyzed these concerns in reports such as *Prospective Research on U.S. Tech Sector: Short-to-Medium-Term AI Computing Trends* (Oct 2025). While debates about the AI bubble remain unresolved, the focus is on identifying the most probable baseline scenario and its market implications.

### Scenario 1 (20% Probability): Substantial Breakthrough in AI Algorithms Despite recent advancements like Google’s Gemini 3, no fundamental breakthroughs have occurred. Challenges persist in scaling laws and multimodal learning, suggesting that major progress may require new technical frameworks. While AI adoption exceeds 60% among S&P 500 firms, applications remain limited to coding and customer service. A breakthrough could trigger a deflationary productivity boom, though this is deemed unlikely.

### Scenario 2 (60% Probability): OpenAI Faces Operational Crisis As a leader in the GenAI wave, OpenAI has intensified the industry’s "arms race." However, Google’s AGI advancements now surpass OpenAI’s, raising risks such as declining ChatGPT subscriptions and weakened fundraising capabilities. If OpenAI struggles, semiconductor and hardware sectors may face earnings revisions, while internet and software firms could benefit. This scenario is seen as the most plausible, potentially fostering a healthier, more pragmatic AI investment environment.

### Scenario 3 (20% Probability): U.S. Inflation Resurgence and Bubble Burst Historical data shows that excessive capital expenditure can lead to resource crowding-out, structural inflation, and liquidity tightening—key drivers of past bubble bursts. With recovering demand in consumer electronics and data centers, AI-related spending risks exacerbating inflation. Coupled with political and monetary policy shifts, a 2026 inflation rebound could mirror 2018’s liquidity crunch, triggering an AI bubble collapse and a brief recession.

### Risk Factors Key risks include unexpected U.S. inflation, OpenAI’s operational instability, geopolitical tensions, slower-than-expected AI advancements, tightening tech regulations, data privacy concerns, and macroeconomic volatility impacting IT spending.

### Investment Recommendation CITIC SEC advises maintaining a cautious "wait-and-see" and "contrarian investment" approach in AI, while gradually increasing exposure to application-side sectors (internet, software). The baseline scenario (OpenAI crisis) is deemed most likely (60%), with algorithmic breakthroughs and inflation-driven collapses each assigned 20% probabilities.