Mission Command—A Way Forward for GenAI by Dr. Timothy Smith

Photo Source: Picryl

A recent report on preliminary findings from research at the MIT Media Lab in Cambridge, Massachusetts, found that 95% of all generative AI projects in industry fail to affect companies' profits and losses (P&L). The report titled "The GenAI Divide: State of AI in Business 2025" looked at 300 publicly disclosed AI initiatives from 52 companies over four sectors of business. Interestingly, the report does not blame AI for the project failures, but instead points the finger at the failure of top-down management. (AI News) These missteps parallel closely to those that lead to tragic losses in World War I. Just as generals ordered devastating frontal assaults from far behind enemy lines, C-level executives, far removed from the real problems facing employees, mandate enterprise AI programs that are disconnected from the real issues impeding innovation and efficiency.

Following WWI, the US Army pursued an evolving change in command structure geared toward decentralized execution, also referred to as Mission Command. Mission Command evolved from observations of leadership failure in the face of a new battlefield forever altered by the invention of the machine guns and barbed wire. In 1916, British General Douglas Haig ordered the disastrous Somme offensive from Château de Beaurepaire, safely located nearly 40 miles from the battle. The battle witnessed over a million casualties for less than five miles of territory. It has become a metaphor for pointless, futile war, and it helped coin the term "chateau generals." Haig's rigid, top-down command structure could not adapt to the reality that machine guns had fundamentally changed warfare. The US Army's shift to a more decentralized approach evolved in direct response to the lessons of World War I and the modernization of warfare. Mission Control refers to a doctrine that emphasizes empowering subordinate leaders to make decisions and exercise disciplined initiative to accomplish a commander's intent. Increasing autonomy became necessary for the "widely dispersed environment" of modern warfare, where lower-level leaders might lose contact with their superiors.

The MIT Media Lab report uncovered interesting elements common to the rare successful generative AI programs that included the necessity for adaptive, learning AI and the grassroots "Shadow AI Economy" as a navigator towards success. The report authors refer to the gap separating successful generative AI projects from failures as the "GenAI Divide." To cross the Divide, the report suggests that companies should invest in adaptive AI that remembers previous interactions with employees and learns from past mistakes, rather than a rigid AI application that will make the same mistakes repeatedly. The authors mentioned how, "A corporate lawyer at a mid-sized firm exemplified this dynamic. Her organization invested $50,000 in a specialized contract analysis tool, yet she consistently defaulted to ChatGPT for drafting work: 'Our purchased AI tool provided rigid summaries with limited customization options. With ChatGPT, I can guide the conversation and iterate until I get exactly what I need.'" This comfort with a tool learned on the side exemplifies the Shadow AI Economy at work. The report details that individuals working in the trenches of each company develop tools tailored to the problems they face. The Shadow AI economy provides a rich source of information on which tools really work and add value to employees' work and productivity.

The story of World War I teaches us that centralized, inflexible planning fails in the face of complex, rapidly changing realities. The move toward Mission Command represented a fundamental rethinking of how the Army operates, transitioning from a rigid, "top-down" bureaucracy toward a system that trusted subordinate units to adapt to evolving battlefield conditions. GenAI initiatives also face complex, changing environments. To combat that daunting 95% failure rate, organizations must embrace a Mission Command mindset as the Army did following WWI. The mindset companies must adopt includes clear intent, not commands from top leaders, autonomy for subordinate leaders to innovate, and careful attention to the Shadow AI Economy for insights into what works best with the new generative AI tools. By unleashing the creativity and domain expertise at the edges of the enterprise, companies can transform AI from a pilot experiment into a sustained engine of productivity and innovation.

Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high-level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

You can buy his book on Amazon in paperback and in kindle format here.