Agentic AI Excitement Outpaces Measurable Results
Investor enthusiasm for agentic artificial intelligence is growing faster than the technology’s ability to generate meaningful revenue and productivity gains, a BCA Research report released on October 23 notes. The report’s chief technology and innovation strategist, Matt Dyson, warned that sustaining the positive sentiment through 2027 will require either stronger AI‑derived revenues or a breakthrough application such as capable personal assistants.
The firm highlighted a persistent gap between AI adoption and measurable financial returns. A 2025 MIT study indicated that 95 % of corporate AI pilots failed to achieve deployment with positive returns within six months. A 2026 McKinsey survey found that only 6 % of respondents attributed at least 5 % of operating earnings to AI adoption.
BCA estimated that AI systems operating with human oversight could perform certain tasks at roughly 75 % lower cost than U.S. employees. However, the firm cautioned that these savings often overlook what it calls the “AI validation tax,” the additional time employees spend checking AI‑generated work.
The report also questioned whether surging AI token consumption accurately reflects productive usage. More sophisticated reasoning models consume four to fifteen times as many tokens, helping explain rapid demand growth without a corresponding increase in useful output.
Agentic AI has raised concerns for traditional software companies. The launch of Anthropic’s Claude Cowork plugins contributed to a 15 % decline in the MSCI U.S. IT Services index in January, affecting companies including Salesforce, Intuit and Workday. BCA described the market reaction as excessive, arguing that smaller businesses are likely to continue favoring established software platforms.
Dyson expects companies will need at least a year to restructure workflows around AI capabilities. He also warned that only 20 % of freelance tasks completed by leading AI models currently meet reasonable client acceptance standards, highlighting the gap between technical performance and commercial usefulness.