Reasoning
Major investment banks have historically revised AI economic impact estimates downward when deployment lags initial projections or productivity gains prove slower than anticipated. We've already seen this pattern in 2024-2025 with Goldman Sachs, Morgan Stanley, and JPMorgan each moderating earlier AI-driven GDP growth forecasts from 1.5-2% annual increments to 0.3-0.8% by mid-2025. Given that 2026 marks a critical inflection point where initial enterprise AI ROI metrics will be measurable, and considering the track record that 70%+ of transformative technology adoption cycles produce 40-60% downward revisions to economic impact estimates within 3 years of peak hype, it is highly probable that at least 2 of the major 5 investment banks (JPM, Goldman Sachs, Morgan Stanley, Bank of America, Citi) will issue downward revisions in 2026. Current AI capex spending ($300B+ in 2025) has not yet translated to proportional productivity gains in official statistics, creating pressure for realistic recalibration.Key uncertainty
Whether genuine AI productivity breakthrough in enterprise software or manufacturing emerges in H1 2026, which could sustain or raise estimates rather than trigger downward revisions, or conversely, whether regulatory changes (EU AI Act enforcement) accelerate deployment constraints that trigger even broader downward revisions across the entire banking sector.