1 Introduction

Images are baked into this HTML file. There is no need to download them separately.

2 AI

The following list identifies the key studies found in the sources that demonstrate the significant challenges companies face in achieving a return on investment (ROI) from AI.

2.1 AI has little or no ROI

1. MIT Study on Generative AI (2025) Key Finding: A staggering 95% of organizations achieved zero measurable ROI despite a combined investment of $30 billion to $40 billion. Only 5% of custom enterprise solutions successfully reached production deployment.
2. Deloitte Global AI ROI Performance Index (2025) Key Finding: Satisfactory ROI typically takes two to four years to materialize, which is significantly longer than the standard 7-12 month expectation for other tech. Only 10% of organizations using "agentic AI" reported significant ROI.
3. Forrester Research: Predictions 2026 Key Finding: Only 15% of AI decision-makers reported an EBITDA lift in the past year, and fewer than one-third can link AI value to profit and loss (P&L) changes.
4. Gartner® Hype Cycle for Supply Chain Strategy (2025) Key Finding: Generative AI has entered the "trough of disillusionment." Fewer than 30% of supply chain AI pilots successfully transition to production systems, and 30% of projects are expected to be abandoned after the proof-of-concept stage.
5. Center for AI Safety / Moon Analysis (2026) Key Finding: AI agents simulating freelance work failed 95% of the time. In a direct comparison, humans generated $150,000 in value from assigned tasks, while AI models managed only $1,700.
6. Goldman Sachs 'AI Narrative Framework' (2025) Key Finding: There is a massive "monetization gap" where the top five cloud providers are spending $381 billion in capital expenditures in 2025, yet only 5% of companies are seeing measurable benefits.

2.2 AI with positive ROI

Top 5 AI Projects with Proven ROI (2026)

1. Fraud Detection & Financial Crime Monitoring
   • Impact: Financial institutions using advanced ML models for real-time transaction monitoring have reported a 2-4x improvement in detection accuracy and a 60% reduction in false alerts, significantly lowering compliance overhead.
   • Source: Master of Code AI ROI Study
2. Supply Chain Demand Forecasting
   • Impact: AI-driven forecasting models continuously analyze logistics, seasonality, and market signals to reduce inventory imbalances, leading to millions in savings through decreased stockouts and waste.
   • Source: Titani Solutions Enterprise AI Examples
3. Intelligent Document Processing (IDP)
   • Impact: Automating the extraction and validation of data from complex invoices and contracts has allowed multinational enterprises to cut processing times by 70%, enabling staff to shift from manual entry to analytical work.
   • Source: IBM Research on AI ROI
4. Predictive Maintenance in Manufacturing
   • Impact: Predictive analytics monitoring equipment health across industrial assets have helped firms reduce unplanned downtime by 20–50%, saving billions in lost production annually.
   • Source: TTMS AI Business Solutions
5. Customer Support Automation
   • Impact: AI-powered agents now resolve routine inquiries instantly across chat and email, with leaders seeing up to 55% automation rates and a 48% increase in response speeds, freeing human agents for complex interactions.
   • Source: NetCom Learning AI Transformation

2.3 AI Bubble

AI Stock Bubble Bursting Predictions

1. Goldman Sachs: Does not predict an AI bubble bursting, but notes that current valuation levels still have a 46% discount compared to the dot-com bubble era.
2. Goldman Sachs: However, they do mention that the Nasdaq 100 Index's price-to-earnings ratio is 46% lower than its peak during the dot-com bubble, indicating that conditions are not yet ripe for a significant market correction.
3. Goldman Sachs: Despite concerns about bubbles, the report states that the current market environment shares some similarities with the late 1990s, but IPO activity is far below the levels seen at that time.
4. Goldman Sachs: The report points out that the rapid expansion of AI workloads will significantly boost data center demand, leading to a projected increase in global electricity demand of over 165% by 2030.
5. Goldman Sachs: Unprecedented investment in AI infrastructure, with the top five cloud service providers’ capital expenditures projected to reach USD 381 billion by 2025, a 68% year-on-year increase, may also be a contributing factor to market concerns.

3 Small business

1. 99.9% of business in the US is small business. According to the latest U.S. Small Business Administration (SBA) 2026 report, small businesses account for 99.9% of all businesses in the United States. SBA.gov.
2. Business Formations: New business applications remain at record highs, with over 1.1 million new formations in the first two months of 2026 alone. Morningstar.com
3. According to the latest U.S. Small Business Administration (SBA) 2026 Frequently Asked Questions, small businesses employ 45.9% of all private sector workers in the U.S.