The Powell Principles of AI
Abstract
Colin Powell’s career stretched from early Cold War duties through Vietnam, Desert Storm, and the digital overhaul of the State Department. His approach to self‑study, team selection, and challenging institutional assumptions offers a useful template for today’s leaders who are integrating artificial intelligence into staff work and mission command. This article argues that commanders should use AI to buy time for critical thought, not to outsource judgement, remaining conscious of AI over‑reliance that can erode critical thinking on the battlefield.
Introduction
“Great leaders are almost always great simplifiers, who can cut through argument, debate and doubt, to offer a solution everybody can understand.” – Retired General Colin Powell
The integration of artificial intelligence (AI) into the global workplace has been hailed as many things: the fourth industrial revolution, an unprecedented leap forward in data analytics, and potentially the end of many white‑collar career paths. Northwestern University researchers suggest that this worst‑case scenario is unlikely. AI is more likely to change how people work than whether people will work at all. Other researchers echo this shift‑in‑work, not end‑of‑work view, even as some labor economists and technologists remain more skeptical about displacement and long‑term disruption. The military and corporate America are no different in this regard. AI will change how many Soldiers perform their jobs and see the battlefield—and a universal consensus on what those changes are is unlikely to be agreed upon by government stakeholders.
In late 2025, the U.S. Army added the 49B Artificial Intelligence/Machine Learning Officer career field, codifying its initiative to capitalize on AI adoption and pursue efficiency‑gaining practices through deliberate implementation. This is not the first time the Army has undergone significant transformational change with its adoption of technology. The Vietnam War marked the Department of Defense’s first attempt to use full‑scale, deliberate commercial computing to pursue an administration’s strategic ends.
During Vietnam, Secretary of Defense Robert McNamara launched a campaign to understand warfare through mass data consumption and quantitative analysis. Friendly casualties, enemy losses, logistics estimates, affiliations of civilian population, and over a hundred other metrics were used by the Department of Defense as key performance indicators to draw correlations from and understand the conflict. Additionally, these statistics were used as a social barometer for the Kennedy and Johnson administrations to interpret the resolve of Vietnamese people and influence public opinion.
Now viewed by many historians and military analysts as an overreliance on performative statistics, the McNamara data initiative was insufficient in its interpretation and lacked clarity in portraying the complexities present in Vietnam. This lesson from Vietnam also shaped a consequential leader for the future application of Army technologies: Colin Powell.
The Powell Principles
The late Retired General Colin Powell offers a compelling example of how leaders can learn new technology, embrace it, and exploit its efficiencies for their organization. After commissioning in 1958 and serving throughout Europe during the Cold War, Powell began his organizational‑level leadership career as a combat advisor in Vietnam in 1962. He sensed the shift in how the Army interpreted battlefield data and, after his second tour, was selected to attend graduate school in 1970.
Powell chose to pursue a master’s degree in computers and information science at George Washington University, preparing himself for future Pentagon assignments amid rapid technological change. He later attributed much of his success as an Army leader to this education and his personal study of how to leverage new technology to achieve results. An understanding of computers and early internet systems laid the groundwork for Powell’s propensity for innovation.
As Chairman of the Joint Chiefs of Staff, Powell consistently pushed for the integration of precision strike, stealth aircraft, and modern command‑and‑control systems into operational planning. His personal comfort with emerging technology did not single‑handedly produce victory in Operation Desert Storm, but it aligned him with broader institutional decisions that emphasized stealth, speed, and overmatch—F‑117 aircraft dominating the air and Abrams tanks outmaneuvering Iraq’s Soviet‑era platforms.
Amid the chaos of the twenty‑first century and the uncertainty of future conflicts, AI poses the promise to be a harbinger of victory if implemented well and to foreshadow defeat for those who fail to master it. How might General Powell have guided leaders today as they adopt AI into their formations? What we do know is Powell left his vision for organizational‑level leadership in The Powell Principles, written by Oren Harari, and that those lessons translate well to the AI revolution. Using The Powell Principles as a lens, leaders can better understand how to inculcate innovative AI usage in their formations by focusing on four areas: setting the example, selecting the team, growing with change, and challenging professional thought. These principles translate cleanly to the AI-era because they deal with how leaders think, learn, and build teams, not with any specific novelty technology or software package. If Powell’s guidance were centered on a particular weapons system or a narrow bureaucratic process, it would have aged poorly. Instead, his focus on personal example, team composition, adaptive execution, and intellectual dissent gives commanders a resilient framework for integrating any new technology, including AI.
Leading from the Front and Setting the Example
Powell invested in his own education long before the Army fully grasped the impact of networked technology. Leaders in the AI‑era face a similar requirement. They must personally embrace a shift in data literacy, prompt writing, and agent engineering. Powell argued that leaders should train for the conditions of tomorrow, not yesterday. Leaders who expect their organizations to become AI‑literate must put in the time and effort to learn its functions themselves.
AI is not another Army program of record—like IPPS‑A, GCSS‑Army, or MEDPROS—where leaders can defer thorough expertise to a few specialists and get by with a cursory understanding alone. AI tools cut across warfighting functions and show up in targeting, logistics, intelligence, and information operations at the same time, which means commanders cannot afford to treat them as niche systems. Multi‑domain warfare is shifting fast, as demonstrated by the Russo‑Ukraine war and current U.S. operations in the Middle East. AI‑enabled tools already shape how commanders view their common operating picture. In Ukraine, commercial satellite imagery, AI‑enabled imagery analysis, and automated targeting tools have helped both sides rapidly identify artillery positions, logistics nodes, and troop concentrations. In recent U.S. operations, automated intelligence analysis and machine‑learning‑assisted pattern recognition have accelerated how joint headquarters process data from ISR platforms into usable targeting information.
Attachment to personal ego and reluctance to change both show up in Powell’s teaching as obstacles leaders must reduce. This personal training time is not about “fighting AI.” It is about preparing to fight an enemy that will use AI against you. Leaders who establish beyond‑baseline AI knowledge for themselves signal to their team the requirement to own this responsibility. AI can quickly translate data into information, and information into knowledge, moving commanders and their staff closer to situational understanding.
Setting the example and maintaining standards is a simple, familiar concept in Army leadership, but it applies directly to AI adoption in warfighting formations. Leaders prioritize what matters. New technology must be integrated like any other unit initiative, not treated as a novelty. Subordinates watch leader behavior closely and often emulate what they see. Leaders who visibly prioritize AI for their own work set conditions for AI tools to enable their teams, rather than let them sit unused in the Army software acquisitions download queue.
Pick the Right People
Powell also offers clear guidance on building teams that can handle technological change. He emphasizes the importance of identifying talent from within the organization. The most promising candidates innately possess intelligence, sound judgement, intuition grounded in experience, loyalty, humility, integrity, and a drive to succeed. Powell’s twist is that he did not reserve these traits for traditional “operators” alone. He deliberately pulled in officers and civilians who could question assumptions, organize between functional areas, network with peers, and translate between technical experts and policy makers.
In practical terms today, Powell would likely tell leaders to seek out bright, motivated Soldiers who have a humble demeanor, a track record of mission accomplishment, and a comfort level working with data and new tools. That is already what any executive officer or deputy commander looks for in building a staff, but Powell’s emphasis on curious, bridge‑building talent will pay extra dividends when forming a coalition of AI‑savvy data analysts and planners. Good team members will recognize their strengths and shortcomings, while leveraging AI to create balanced solutions to problems.
What does this look like for AI implementation in a brigade or division staff? During the Military Decision‑Making Process, leaders begin by alerting the staff and gathering the tools. Organizational‑level leaders know who is coming to the table, but they may not know every tool those people can bring. Leaders should update their estimate of their team’s talents and deliberately encourage experimentation with AI tools.
Challenges, goals, and small tests can push the team to explore the capabilities of AI and reveal which members have an instinct for prompt writing, vibe coding, writing instructional scripts, creating workflows, or developing agents. When leaders do this, they are not just assigning a science‑fair project. They are discovering who on the staff has an innate feel for how AI can help the mission.
Once identified, these individuals can advise on where to bring AI into staff processes. They can help build a simple accountability framework for AI use, identify which products make sense for AI support, and point out where human staff work should remain untouched. Over time, this AI‑literate cadre becomes a keystone for integrating AI into the operations process without allowing the technology to dominate it.
Let Situation Dictate Strategy
Powell argues that effective leaders must grow with changes in their environment rather than cling to the methods that worked during their last assignment. Organizational‑level leaders need to keep enough latitude to adjust their execution and change strategy as conditions evolve. For AI, this translates to incorporating AI into the battle rhythm while preserving the ability to switch back to conventional information analysis as the fog of war clears or when systems fail.
Commanders set conditions so their staff can work swiftly with all available tools and still have time to move through the cognitive hierarchy from data to information, to knowledge, and ultimately to understanding. AI has the potential to weaponize the speed of data processing and information analysis. If applied well, it can buy time—time for humans to think critically, challenge assumptions, and refine recommendations before they reach the commander.
Leaders in the Profession of Arms understand both sides of the risk: the danger of lagging situational understanding and the danger of hasty error. Given that tension, the answer is not “AI or no AI.” Leaders must do both and seek balance. Experience using AI enables leaders to understand where to accelerate parts of the problem with computation and where maintaining human judgement is invaluable. Balance and adaptation in AI implementation mitigate inherent risks.
AI analysis of mission command inputs can provide leaders and staff with an objective view of how information connects and correlates. This knowledge can quickly generate initial courses of action, offering commanders flexibility and speed. Powell, however, would remind leaders to stay vigilant when interpreting information and to push their thinking beyond their comfort zone. Stay balanced, stay rooted, but expand your idea of what is possible.
Modern leaders seeking to empower their staff should view AI as a tool to automate the ordinary. At the same time, they must develop teams that are not over‑reliant on AI agents. Mastery of both AI‑augmented and traditional methods guards against regression in critical thinking skills and preserves the mental power needed to perform staff functions under stress.
Don’t Be Afraid to Challenge the Pros
Powell’s approach to organizational leadership included a willingness to challenge the thinking of his professional peers. Today, many organizations adopting AI assume that AI automatically reduces monotonous workload and buys back time for higher‑value work. Powell would likely question that claim.
He would probably argue that everyone in the organization has a responsibility to challenge universally held beliefs that rest only on positive metrics and vendor promises. As with any innovation, growing pains emerge as technology moves from beta, to testing, to full rollout. AI is no exception.
Recent civilian workforce studies highlight a cycle that can appear in AI adoption. AI tools help users complete assigned tasks faster. Managers see the increased throughput, assign more work, and normalize higher expectations. Other research on automation and digital monitoring in offices and warehouses shows similar drift: once a new tool boosts productivity, organizations rarely use the gain to create slack. Instead, they ratchet up output targets and expand the volume of tasks. In that environment, team members become progressively reliant on AI as their responsibilities grow. Over time, subordinates feel pressure to move faster. Their judgement and reasoning suffer, and they struggle to spot errors in AI‑generated outputs.
Viewed through Powell’s lens, this cycle should trigger concern. As noted by Powell, stress, reduced situational awareness, and fatigue are major contributors to mistakes. By challenging the assumption that AI can only help, leaders can counter confirmation bias and open space for honest discussion up and down the chain.
In doing so, they help their teams work at a sustainable pace and maintain the cognitive space to identify errors before those errors become catastrophic. The Army differs from corporate America in this instance. In civilian business, a mistake might mean a customer order is billed to the wrong account or that a sales presentation is poorly received. In Army leadership, mistakes carry a human cost. This is a distinction that General Powell communicated regularly to his team—hapless haste gets people killed.
Training for Tomorrow
The Powell Principles enable practitioners to lead teams that integrate AI into staff processes to gain a tactical edge. By setting a personal example, selecting a team of AI‑capable experts, growing with change, and challenging professional thought, leaders can bring AI into their organizations in a deliberate way that supports the operations process rather than distorting it.
By creating room for AI experimentation, leaders facilitate the compounding innovation and opportunities required for victory in large‑scale combat operations. To weaponize AI without becoming dependent on it, leaders must both encourage innovation and use of AI, while codifying when human judgement means the most. Powell’s career was full of victories he partly attributed to embracing new technology and using its strengths, while remaining skeptical of its limitations. By studying his example, today’s leaders can interpret where AI can viably benefit their organization and the Army as a whole.
Disclaimer: The views expressed are those of the author and do not reflect the official policy or position of the Michigan Department of Military and Veterans Affairs, the Michigan National Guard, the US Army, Department of War, or the US Government.
