For more than a year, financial markets have eagerly valued the artificial intelligence (AI) surge as a purely positive, “upside-only” development for asset prices, a view largely shared by typically optimistic analysts at Research. However, on Friday, the bank’s more defensively positioned European equity strategy team cautioned that this uninterrupted euphoria is set to fracture by early 2026.
In a client note reviewed by , Bank of America strategists stated that “doubts around the AI revolution are emerging,” as the market story quickly changes from an “upside-only” outlook to grave worries that AI represents a “double-edged sword.” A primary new concern is the increasing awareness that AI may not broadly enhance corporate earnings—it could actively erode them.
BofA emphasized several significant “downside risks” that are, frankly, dampening enthusiasm for the AI trade. Traders now face a landscape featuring cannibalization, capital expenditure reductions, and other assorted AI-related challenges.
The Threat of Profit Cannibalization
BofA identifies a major oversight in prevailing market expectations. The sellside consensus currently forecasts an impressive 17% compound annualized earnings per share (EPS) growth for global equities over the coming five years. This sits at the core of what strategists term the “cannibalization” paradox.
Corporate profit margins are already at record levels, BofA strategists observed, making this 17% projection difficult. Historically, double-digit compound EPS growth has only occurred when margins were low. These “optimistic” forecasts therefore presume that tech EPS growth driven by AI adoption will be lasting and, simultaneously, that this expansion will not “cannibalize existing profit pools.” This is precisely where investors are undergoing “the largest rethink,” evidenced by steep declines in AI-sensitive sectors like software, insurance, and wealth management, and a shift toward safer sectors dependent on physical assets (e.g., mining, utilities, and chemicals).
This harsh reality is already impacting the market. For example, , an early software leader traded as an AI beneficiary, recently saw its substantial rally suddenly reverse. The plunge in software-as-a-service (SaaS) stocks, which lost over $2 trillion in roughly two weeks, was consequently labeled the “SaaSpocalypse.”
Labor Market Fragility and Capex Cracks
Beyond direct business cannibalization, BofA warned that the AI rollout carries serious macroeconomic dangers. The U.S. labor market is already showing cautionary signals, with three-month payroll growth stuck at a weak 0.1%—a rate of virtually no job growth historically linked to the conclusion of previous equity bull markets. This includes both 2000 and 2007, when bursting bubbles aligned with severe recessions.
BofA alerts that AI-driven productivity improvements could further reduce corporate demand for workers, intensifying this softness. Paradoxically, substantial job losses might harm the tech giants themselves, since the hyperscalers financing the AI boom depend significantly on consumer advertising income, which requires a robust, employed consumer base. Analysts contend that the underlying assumption in consensus market and earnings forecasts is that labor market weakness stems from temporary factors like trade conditions, immigration restrictions, and over-hiring in the post-pandemic “Great Resignation” period, but what if these assumptions are incorrect?
Moreover, the debt-driven AI investment frenzy is displaying clear signs of stress. For the first time in the history of BofA’s Global Fund Manager Survey, investors indicated they believe companies are investing too much and should reduce capital expenditures (capex). Corporate bond spreads for U.S. hyperscalers have reached a three-year peak, and private credit platforms with heavy exposure to software debt are faltering under strain, exemplified by Blue Owl halting retail investor redemptions for one of its funds this week. BofA’s U.S. equity strategy team anticipates these challenges will peak in an “AI air pocket” later in 2026.
Finding “AI Hiding Places”
Compounding the pessimism, BofA warns that the anticipated economic lift from AI might not appear. While equity markets are currently factoring in nearly 3% U.S. productivity growth, forecasters like the Congressional Budget Office estimate AI will increase productivity by a mere 0.1% per year over the next decade. In line with this, new data on Friday showed the in Q4 2025 was just 1.4%, following stronger-than-expected growth for most of the year.
In reaction to this evolving environment, BofA warned that AI-infrastructure providers—such as semiconductors, capital goods, and construction materials—now appear dangerously “stretched,” trading at record-high relative valuations and peak earnings expectations. Deeming them highly susceptible to AI capital expenditure letdowns, BofA has sharply downgraded the semiconductor sector to an underweight rating.
Among hyperscalers, fissures are becoming discernible amid in late 2026 as funding gaps surface for its immense computing needs. CEO Jensen Huang that his pact to invest $100 billion in OpenAI was “never a commitment” while maintaining he still fully valued his partnership with Sam Altman’s firm. Owen Lamont, a portfolio manager at Acadian Asset Management and a former University of Chicago finance professor, that the current market situation meets three of his four custom criteria for a bubble, lacking only a surge in IPOs.
Rather than pursuing the AI rally, BofA recommends investors look for cover in “AI hiding places.” The firm is holding overweight stances in defensive sectors with minimal AI disruption risk, like consumer staples, telecoms, and chemicals. Overall, the bank maintains a firmly negative stance on European equities, predicting a 15% decline by Q2 as the difficult reality of a cannibalized, double-edged AI market takes full hold.