¿El sector de IA está entrando en una corrección o en una burbuja?

The artificial intelligence sector, once a beacon of unbridled optimism, now faces mounting scrutiny as investor skepticism intensifies. By late 2025, the market's initial euphoria over AI's transformative potential has given way to a recalibration of valuations, driven by concerns over capital efficiency, profitability, and the sustainability of speculative growth narratives. This shift raises a critical question: Is the AI sector entering a correction-a temporary realignment-or a full-blown bubble reminiscent of the dot-com era?

The Current Valuation Landscape: Optimism Meets Caution

The AI sector's meteoric rise between 2024 and mid-2025 was fueled by unprecedented demand for AI infrastructure, with companies like NVIDIA and Oracle dominating headlines. However, recent developments have cast a shadow over this optimism. Oracle's delayed data center projects and unexpectedly high capital expenditures, for instance, have triggered a broader selloff, with investors questioning whether the sector's growth is underpinned by tangible returns or speculative hype according to market analysis. Bank of America's October 2025 survey of global fund managers further underscores this unease, revealing that 54% believe AI stocks are in a bubble.

Despite these concerns, the sector retains strong fundamentals. Leading AI firms such as Microsoft, Alphabet, and NVIDIA have demonstrated robust revenue growth and profitability, distinguishing them from the largely unprofitable dot-com companies of 2000. For example, NVIDIA's earnings per share (EPS) surged from $0.72 in May 2023 to $2.34 by December 2025, while OpenAI's annual revenue reached $13 billion according to financial analysis. These metrics suggest a market grounded in real economic value, albeit with elevated valuations.

Historical Parallels: Dot-Com vs. AI

Comparisons to the dot-com bubble are inevitable, but key differences emerge when analyzing financial fundamentals. During the dot-com peak, only 14% of internet companies were profitable, and the Nasdaq-100's forward P/E ratio hit 60x in March 2000. In contrast, the Nasdaq-100's forward P/E in late 2025 stands at 26x, a significant discount to historical extremes. Moreover, today's AI sector is characterized by higher enterprise adoption rates, with 87% of large enterprises implementing AI solutions by 2025. This contrasts sharply with the dot-com era, where many internet-based businesses lacked proven business models and broad adoption.

Yet, parallels persist. Circular financing structures, such as NVIDIA's $100 billion investment in OpenAI in exchange for GPU purchases, mirror the telecom capacity swaps of the late 1990s, where companies inflated revenue through self-reinforcing deals according to industry analysis. While these arrangements provide short-term stability, they raise concerns about long-term sustainability. Additionally, private AI startups command valuations 56% higher at Series C and 230% higher at Series D+ than non-AI peers according to market research, signaling potential overvaluation in the venture capital space.

Investor Sentiment and Capital Rotation

Investor sentiment toward AI stocks in late 2025 reflects a nuanced duality. While capital continues to flow into AI-native companies with clear paths to profitability-such as those offering AI-driven customer-facing applications-there is a growing preference for mid-term financial discipline over speculative long-term bets according to market insights. Private equity firms, for instance, are increasingly targeting AI opportunities that deliver measurable cost efficiencies and predictable returns, prioritizing pragmatic value generation over hype.

This shift has triggered a market rotation out of high-valuation tech stocks and into value-oriented sectors like financials, energy, and industrials according to market analysis. The selloff in companies like NVIDIA and Oracle underscores this trend, as investors seek safer havens amid uncertainty. However, the AI sector's resilience-bolstered by enterprise adoption and real revenue growth-suggests that a full collapse is unlikely. Instead, the market may be entering a phase of selective correction, where overvalued subsectors face downward pressure while fundamentally sound companies retain their appeal.

The Path Forward: Balancing Growth and Prudence

The AI sector's trajectory hinges on its ability to balance innovation with financial discipline. While current valuations are elevated, they remain well below the extremes of the dot-com bubble, supported by tangible technological adoption and scalable infrastructure. For instance, 70-78% of global companies now use some form of AI, indicating a structural shift rather than a speculative frenzy.

However, risks persist. The sector's reliance on circular financing and the potential for idle data center capacity could undermine the "insatiable demand" narrative according to market analysis. Investors must also contend with the Shiller P/E ratio for the S&P 500 exceeding 40, a metric historically associated with overvaluation according to economic analysis.

Conclusion: A Correction, Not a Bubble

The AI sector is not in a bubble akin to the dot-com era, but it is undeniably in a correction phase. The market's re-evaluation of valuations reflects a maturing investor approach, where growth is tempered by demands for profitability and capital efficiency. While speculative excesses exist-particularly in private markets-the sector's strong fundamentals, including enterprise adoption and established profit-generating models, provide a buffer against a catastrophic collapse.

For investors, the key lies in discerning between AI companies with sustainable business models and those relying on hype. As the sector navigates this recalibration, those prioritizing real value creation over short-term speculation are likely to emerge unscathed.