How World War II Bureaucratic Sabotage Endures in the Defense Department, and How to Fight Back – Texas National Security Review

During World War II, the Office of Strategic Services — predecessor to the Central Intelligence Agency — produced a manual detailing methods of sabotage intended for the European Allied resistance in German-occupied areas to disrupt the German war machine. But this wasn’t about blowing up bridges.1 The Simple Sabotage Field Manual established disruption tactics to impede productivity and create inefficiencies in enemy organizations. These tactics were considered subversive acts of warfare.

Ironically, from rigid adherence to procedure to endless deliberation in committees, many of these tactics have become normalized and even embraced in modern workplaces. How did we come to unwittingly replicate this wartime sabotage directed at our enemies to proliferating it against ourselves?

I think clarity can be found in our relationship with time and risk. In times of peace and stability, we perceive we have more time to fill, our relationship with risk drops, and we make people do more things to ensure less risk. Then if war, conflict, or emergency hits, we abandon those same sabotage and bureaucratic tactics because “things really matter” and a lazy game becomes speed chess, on the clock.

What if we behaved as if time was always the most powerful weapons platform we have and every minute mattered, every day? What if we respected people’s time, lessened the toil, and asked, “Would I do this in war?” What if we treated our daily work as the “practice field” for war, rather than only scrambling to overcome all the malaise when it becomes “game critical”?

In the face of a new technological revolution, these bureaucratic tendencies are especially obstructive as they stifle the adoption of technologies like AI — a critical technology needed to maintain and sustain American power. With its ability to streamline processes and cut through complexity, AI offers the promise of eliminating the very inefficiencies the Simple Sabotage Field Manual was designed to create. Yet, the entrenched practices of over-caution, rigid adherence to channels, and endless committee reviews mirror sabotage in a way that not only slows progress but actively prevents AI’s transformative potential from being realized.

Counteracting Modern Sabotage Practices in Government Matters Now

One of the key reasons to prioritize stopping organizational sabotage is because our mission really matters. In national security, seconds matter. Being able to move quickly and confidently matters. Being able to ask and answer “What if?” faster than the adversary matters. But most importantly, we, the people, matter. The worst part of these sabotage tactics is they slowly bleed the purpose out of people. These tactics ultimately translate into toil, and toil kills purpose faster than mission can replace it. At a time when recruitment, retention, and civic interest are at an all-time low, we cannot afford to suck the soul out of the people we have. We should unleash, empower, and inspire our teams by tackling the toil, rejecting the sabotage, and demanding more for ourselves, our mission, and our nation.

One of the most interesting tensions we face as leaders in government is that as we rise in rank, the toil is ironically removed from us, and we are often the condoning and requesting sources of sabotage. The sabotage is OUR signatures, OUR meetings — WE the leaders are the saboteurs.

This can lead to a dangerous lack of empathy for what our people and our teams must navigate every day just to do their jobs. We say, “It’s just a form,” “It’s just a minute,” or “It’s just one training.” The danger of “it’s just” is that it shows our lack of respect for our people, their purpose, their passion, and their time. I have a wristband that I wear with the acronym “WIDTIW,” which stands for “Would I Do This in War?” It reminds me to prioritize the most important things and to protect minutes on mission.

As public servants, it’s time to take a hard look at how we approach our daily business.

Meanwhile, our people slowly bleed their sense of purpose on each form, tick box, and meeting. We slowly beat the curiosity out of them, draining their energy and attention when we need to reinvigorate and inspire them. Newer generations in the workplace understand exactly what purposeful work is versus what toil is, and they do not want to be sabotaged. We can complain about Generation Z, but maybe they are teaching us one of the most critical lessons that we can learn: Time is precious. Time is an advantage. A minute is as critical as a missile. We can’t make more time, we can’t buy more time, we can’t trade for it, and we can’t contract for it — when it’s gone, it’s gone, and so is our national security advantage.

But not all hope is lost — recognizing and addressing these counterproductive behaviors, this sabotage could be key to modernizing and streamlining government operations today. Each sabotage tactic outlined in the Simple Sabotage Field Manual has a corresponding set of countermeasures that can — and should — be carefully considered and implemented. As public servants, it’s time to take a hard look at how we approach our daily business. Being open to new tools, including AI, will be critical to cutting toil and giving us minutes on mission.

Tactic 1. Insist on Channels: The Tyranny of Bureaucracy

The Simple Sabotage Field Manual first advises saboteurs to harness the tyranny of bureaucracy and insist on doing everything through official channels, avoiding shortcuts that might expedite or more quickly inform decisions.2 Insistence on following strict protocols and procedures is commonplace in today’s organizations. Employees are often required to navigate a labyrinth of approvals, forms, and signatures before they are permitted to act. This adherence and allegiance to process over practicality and efficiency, especially in a digital age, slows decision-making to a crawl and disrupts productivity. For example, when employees request specific software, a five-minute task can turn into weeks-long ordeals, requiring approval from multiple departments. Similarly, access to many government systems require employees to complete online paperwork that, ironically, can only be accessed with a common access card or personal identity verification card whose critical feature is to identify and validate the user’s right to access those same government materials and systems.

Countermeasure 1. Cut the Red Tape: Talk Directly, Act Swiftly

Instead of remaining wedded to highly bureaucratic decision-making processes, organizations should delegate authority to frontline employees and managers, allowing them to make decisions without needing to go through multiple layers of approval, in all but exceptional cases. At the Ritz-Carlton, every employee is authorized to solve a problem or create a delightful guest experience of a value up to $2,000 irrespective of position: maid, groundskeeper, server, etc. What would it look like to empower our people to solve the issue where they see it, when they see it?

Consider introducing “fast-track” programs where certain decisions can be made more directly and quickly. If successful, these programs can be expanded to other areas, reducing reliance on channels.

Some organizations have created “Risk Appetite Statements” to help employees understand when the organization wants them to take healthy and considered risk and when they do not. This helps give clear “commander’s intent” to each person.

Removing channels is also about being willing to delegate authority and disentangle our sense of identity and value from being a gatekeeper versus a direct doer. The more “doers” we enable versus critics or gatekeepers, the faster the solutions are implemented, and the happier everyone tends to be.

Tactic 2. Refer to Committees: The Endless Cycle of Deliberation

The manual advised that all matters should be referred to large committees, ideally with five or more members, ensuring endless deliberation and no decisive action.3 This cycle of deliberation without decision is a hallmark of modern governance and one of the more pervasive sabotage tactics.

Government often relies heavily on committees, boards, and task forces to reflect on and deliberate issues, sometimes over many months or even years. The larger the committee, the less effective it is. We have seen this recently in attempts in almost every government organization to form committees to address AI use. After months of meetings, these groups often fail to produce actionable or meaningful solutions or simple functional policy and guidance, or else the outputs are so over-generalized that they provide nothing truly useful to those who are practically navigating the technology.

In national security, the adversary couldn’t have designed our approach to AI to better achieve sabotage. Though the United States maintains global dominance over AI technology (for now), my grandmother has easier access to advanced technology than most security-cleared public servants.

Countermeasure 2. Skip the Committee Carousel: Decide Without the Dance

Consider the impact of the committee not coming to an actionable conclusion within a set period. If there is no real impact on the actual work happening, there is likely no need for a committee.

When forming a committee, provide a clear and narrow mandate with specific goals, deadlines, and decision-making authority. Keep committees small and composed of relevant experts. Reward committees that produce actionable recommendations within set timeframes, and hold committees accountable for delays, including disbanding them early and often when they are not making progress.

Tactic 3. Haggle Over Wording: The Paralysis of Precision

The manual recommended arguing over precise wording, a tactic designed to delay and frustrate all involved.4

Most importantly, in national security, the adversary will continue to advance, regardless of whether we agree on definitions or not.

When drafting modern government policies, legislation, or even internal memos at all levels, significant time is often spent debating exact language. This focus on minutiae delays important initiatives. For example, at an event just the other night, someone asked the speaker what the definition of AI was. The speaker’s response was “there is no agreed-upon definition” — and yet we have somehow created thousands of global policies, doom-looped the narrative, and established what is likely thousands of committees, boards, and oversight groups in different organizations, all while hundreds of millions of people and our adversaries are using and have been using AI in various forms for decades.

Countermeasure 3. Get the Gist: Skip the Semantics

Teams that experience frequent delays related to semantics should be encouraged to focus on the intent and substance of communications. Emphasize that perfect wording is less important than timely, clear communication. If these efforts prove unsuccessful, consider designating a final arbiter to settle definitional disputes quickly, preventing prolonged discussions.

Put wording into context. If a lack of agreement on a definition has no impact on whether a topic will continue or advance, then it is easy to recognize when the haggling is sabotage. For example, innovation continues regardless of the decades of debate on its definition, as will AI and other advancing technology. Most importantly, in national security, the adversary will continue to advance, regardless of whether we agree on definitions or not. Our job is to outpace the adversary, not to out-bureaucratize them.

Tactic 4. Demand Written Orders: The Bureaucratic Demand for Documentation

Saboteurs were also told to insist on written documentation for all orders, ensuring that nothing in an organization could be done quickly or informally, or rely on trust and assumed competence.5

Despite the proliferation of effective and secure digital technologies, the government remains heavily reliant on physical documentation. Requests, approvals, and directives often require written confirmation as each manager, concerned with liability, demands extensive documentation before signing off. Urgent projects are delayed because key decisions require written approval from several levels of management. In other cases, multiple forms must be completed to access data.

We say we want trust and psychological safety in our organizations, yet we refuse to trust. Documentation is often a physical manifestation of “cover your ass.” While it makes sense to document some things, the reality is most things of importance in government happen between people who trust and respect each other. When we try to substitute tick boxes, forms, and paperwork for that trust and relationship, we remove the “coming together to solve the problem” experience that creates shared knowledge, values, and clarity of mission and priorities. We wonder why our organizational cultures are failing as we try to substitute culture with paperwork and are surprised when people leave our organizations. The reason most people stay at an organization is because they like and feel connected to the people they work with. When we remove the “peopling,” the perceived value of the work quickly falls to the lowest common denominator of a paycheck.

As an authorizing official, one of the most critical things I can have is a strong relationship with other authorizing officials and cyber professionals. For a technology to scale its adoption and application, reciprocity is key, yet reciprocity ultimately relies on trust.

Ironically, while many people don’t trust government, the media, their neighbors, or each other, they want their AI to be “trustworthy.” What have we become when we expect the machines we create to be better ideals of trustworthiness than ourselves and each other?

Countermeasure 4. Good Enough Is Perfect: Precision without Paralysis

While documentation is important, allow verbal approvals for non-critical tasks, with written documentation to follow, if legally necessary. This strategy can speed up decision-making while maintaining a paper trail for accountability. Early on, set clear expectations about when written orders are necessary, and make them as simple as possible.

Begin using digital tools and technology to streamline the process of issuing and receiving written orders and automating documentation processes.

A great tool is the “Assume Go” memo, where instead of the person championing the change having to exhaustively outline why the change is good, those against the idea must write the case on why the change would be detrimental. If you make it harder on the person who is trying to do something new and good than you make it on the person getting in the way of change, unsurprisingly you will get less change. When you force someone through a 63-field form to try basic generative AI on public information, you make them emotionally daunted before they type information into the first field. Flip that by asking those who say they shouldn’t use AI to fill out the form to explain why the AI use case is a bad idea.

Begin using digital tools and technology to streamline the process of issuing and receiving written orders and automating documentation processes. In many digital collaboration tools, changes and edits by different users are tracked, removing the requirement to formalize those types of alterations. At the U.S. Agency for International Development and Google, I could open a document and simply initial when I was done reviewing it, thus generating a permanent record. This process required no PDF generation or cumbersome printing and sending and was always available for reference if people had questions.

Tactic 5. Make Speeches and Meeting Mania: Talking Without Saying or Doing Anything

The Simple Sabotage Field Manual suggested holding conferences or frequent meetings when there is more critical work to be done, distracting from essential tasks.6 It encouraged saboteurs to talk frequently and at great length, filling speeches with long anecdotes and irrelevant details.

Readers may recall meetings that devolve into marathons where members recount their experiences and opinions on unrelated matters and add “colorful” personal stories, leaving key agenda items unaddressed and decisions unmade. Elon Musk famously gave his teams permission to just walk out of meetings they weren’t adding to or getting value from.7 Can you imagine the response if we did the same in government?

The modern workplace is plagued by an overabundance of meetings. These meetings often run long, resulting in employees spending more time discussing work than doing it.

In a 2017 survey, 65 percent of managers said that meetings hold them back from completing their own work, 71 percent said meetings are unproductive and inefficient, and 64 percent said they had to sacrifice deep thinking and time to be strategic for those meetings.8 Data shows that managers spend 50 percent or more of their work week in meetings, while 70 percent of employees feel their satisfaction would improve if they attended fewer meetings.9 Employees spend 31 hours or more in unproductive meetings each month, and 76 percent of employees prefer to keep the start and end of the week meeting-free. A study found that video meetings are even more draining to people because of excessive required eye contact, watching yourself and being anxious about your own reactions for prolonged periods, being stuck in one place for too long, and high cognitive load.10

Slack’s Workforce Index study found that, for most people, two hours of meetings a day is the most they can stand without impacting their happiness, which translates to 25 percent of the work week.11 So if we want to retain great talent in government, we need to consider the impact of meeting toil in their likelihood to stay public servants.

Countermeasure 5. Ditch the Drivel: Say What Matters and Gather with Purpose

Ensure every meeting has a clear, concise agenda with specific objectives outlined at the top.

Make meetings shorter. Parkinson’s Law states that people tend to fill whatever time is allotted for a task, and team members will consciously or subconsciously fill 30- or 60-minute meetings even if it takes less time to achieve the goals. Try starting with 15 to 25 minutes.

Be ruthlessly aggressive about canceling meetings when emails and other information-sharing can supplement.

Limit the time allocated for each topic to prevent long-winded discussions, and encourage participants to stay on-topic. Consider introducing a time limit for individual contributions during meetings. Some productivity tools have “speaker metrics” on their video call technology so you can see who is speaking (or maybe over-speaking) in the meeting. If the meeting has members who are not impacted by or impacting the issues, allow them to leave. Keep required meeting participant numbers as small as possible. Meetings are not the only way to make people feel included in a process, decision, or work.

Be ruthlessly aggressive about canceling meetings when emails and other information-sharing can supplement. Clear the calendar of all recurring meetings twice a year, and prohibit reoccurring meetings from being scheduled for at least two weeks. Teams often find that having reoccurring meetings on the calendar can slow the progress of a project to the cadence of the scheduled meetings, not the pace in which the project can actually be achieved. Set “no meeting” days so that people can get into “flow state” and better focus on deliverables and progress without interruption.

Constrain the size of preparatory documents to only a few pages. Amazon famously has a six-page limit and gives people time at the beginning of the meeting to read the paper silently before discussion, which prioritizes clear communication over flashy presentations.

Outline if a meeting is for decisional purposes and who is empowered to make the decision. Focus meetings on outcomes, shifting the emphasis from discussion to decision-required. Require that each meeting ends with specific action items, assigned responsibilities, and deadlines, ensuring that discussions lead to tangible results.

Tactic 6. Apply Regulations Rigidly: The Inflexibility of Compliance

Saboteurs were advised to apply all rules and regulations to the letter and add new regulations wherever possible, ensuring there is no room for flexibility or common sense.12

Ironically, as employees become more trained, they are not necessarily more trusted. Employees are often required to adhere strictly to extensive rules and regulations, at the expense of efficiency, practicality, quick problem-solving, intuition, and innovation. An emergency request for resources may be delayed because regulations dictate that the request be denied until the proper paperwork is submitted — sabotage success!

One of the most critical factors to efficiency is empathy and respect. Unfortunately, many approach their work without empathy, as administrators who assume that every worker is a naughty child who requires constant supervision. Much like a left-lane or speed-lane “sitter” who goes at or below the speed limit, these individuals feel empowered to slow down and stop others with righteous indignation. Ironically, many states have now created laws and tickets for these so-called “lane sitters” to underscore that the left lane is for passing. What if we, too, assumed that public servants are not trying to be intentionally nefarious — that, in fact, they truly care about the mission and are simply trying to get things done? What if we ask, “How do we get out of a great public servant’s way”?

Countermeasure 6. Bend, Don’t Break: Smart Compliance

Give managers the authority to waive or adapt regulations in situations when rigid application would be counterproductive and train them to use discretion and common sense. Part of this is culture, and intentionally hiring people who want to manage versus administrate. Managers want to be given a goal and a set of resources to accomplish that goal. Administrators want to be given the rules and to enforce them. When I was the telework executive at a federal agency, I saw “manager-minded” people who knew how they wanted flexibility to look across their teams and their work and would use their discretion to deal with high and low performers, using telework as a tool to increase performance, progress, and satisfaction. “Administrator-minded” people were uncomfortable with that discretion and flexibility because it required them to make tough calls and have tough conversations. Instead, they wanted to make rules like “everyone will telework one day a week”; they wanted to hide behind the guise of “fairness” versus having to manage performance. Similarly, if they had one problem person, instead of removing the telework of that person, they would say that “telework isn’t working” and take it away from everyone. Not surprisingly, high performers are quick to abandon teams administrated in this way.

Policy complexity in the beginning exists because there is often an “unknown” element to the actual risk the policy is intended to navigate, or it assumes that everyone will behave as the aforementioned naughty child instead of a trusted, ethical, and mature public servant.

Review regulations to identify those that are outdated or overly rigid. Simplify or eliminate those rules that no longer serve their intended purpose, making compliance more practical. A policy shop’s goal should be that every few months, a policy should get shorter and simpler. Policy complexity in the beginning exists because there is often an “unknown” element to the actual risk the policy is intended to navigate, or it assumes that everyone will behave as the aforementioned naughty child instead of a trusted, ethical, and mature public servant. As public servants do engage and the risk does not materialize, this evidence should be considered, and policy made more streamlined to reflect the actual behavior of people versus the feared behavior.

Generative AI tools are great examples of this. Many leaders ask, “But what if people misuse it?” The bottom line is that we should penalize those who do, but the majority are simply using AI to reduce toil and get minutes back on mission. We shouldn’t be quick to criminalize assumed behavior. Instead, let’s be intentional in our “learning by doing” and reward good behavior, impact, and returns.

Tactic 7. Bring Up Irrelevant Issues: The Art of Distraction

The manual advised saboteurs to distract meetings and discussions by bringing up irrelevant issues.13

It is commonplace for discussions to drift into tangential topics often unrelated to the core agenda. These distractions can consume valuable time and prevent critical decisions from being made. For example, a meeting focused on budget allocation may be sidetracked by a lengthy debate on data governance. While the discussion may be valid, it is irrelevant to the primary issue, resulting in delays and unfinished business.

Countermeasure 7. Stay on Target: No Squirrel-Chasing

Establish an agenda and circulate a pre-brief that outlines the key topics, objectives, and any preparatory work needed, and who is a presenter, participant, or decision-maker.

Consider using a “parking lot” strategy, where off-topic issues that arise during the meeting are immediately noted and set aside for later discussion. This keeps the current meeting focused while ensuring that other concerns are acknowledged. Assign a “parking lot attendant” or facilitator to gently steer conversations back to the agenda when distractions arise.

Tactic 8. Reopen Decided Matters: The Never-Ending Debate

Saboteurs were instructed to refer to previously decided matters, reopening discussions and questioning decisions that had already been made.14

Regardless of whether this phenomenon is due to new information, changing political climates, or simply indecision, reopening settled matters is a common occurrence in modern-day government that stalls progress, creates a sense of instability, and undermines action. A group directed to explore the relevance of AI by doing a small pilot may be empowered in one meeting, only for the decision to be rescinded in the next meeting. The constant revisiting of decisions effectively causes delays and erodes trust in the people and the process. More importantly, it erodes the sense of curiosity, excitement, and passion that a public servant might have for their work and their mission.

Countermeasure 8. Decide and Drive: Skip the Second-Guessing

To avoid debate on settled matters, clarify who the decision-makers are and who gets to reopen previously closed items. Holding individuals accountable for reopening decisions without valid cause can also limit excessive debate. Individuals who seek to reopen decisions should be required to present a thorough justification and/or analysis to reopen a matter. It should never be easier to be a critic than it is to be a champion or a doer — if it is, your organization will drown in critics and starve for doers.

Consider setting a “default decision” upfront — lay out an assumed decision in a specific timeframe, and if nothing of significance counters that decision during that timeframe, then it should go into effect on the date allotted. For example: “The Policy of allowing the use of Generative AI by our staff for publicly available information will go into effect on Dec. 15, with the requirement to sign the acceptable use clause and take the provided AI training.” Any dissent, with evidence and required risk mitigation measures, will be accepted for consideration prior to Dec. 10. If no credible dissent or additional measures are identified, the policy goes into effect on Dec. 15.

Tactic 9. Advocate Caution: The Paralyzing Power of Paranoia

The Simple Sabotage Field Manual recommended saboteurs advocate caution in workplace decision-making, urging colleagues to avoid haste to prevent possible embarrassments or difficulties.15

Government officials may overemphasize caution to avoid perceived risk, without realizing that doing nothing can be the riskiest choice. While caution is necessary in some circumstances, excessive focus on it can lead to paralysis by analysis. Many government agencies spend months or years analyzing the potential risks of implementing a new program. The constant calls for review, external assessment, and caution prevent the program from ever getting off the ground, or, in many cases, eat up so much of the budget that there isn’t enough money left at the end to do the work.

Productive and intentional critics like lawyers already exist to partner with government innovators to appropriately explore and suggest ways to mitigate against harm.

In more extreme examples of this sabotage tactic, some government agencies mandate the introduction of new “critic-based” roles. A recent example of this phenomenon is the introduction of “AI ethicist” roles, which, by their nature, infer that government employees are not already expected to behave morally, ethically, and thoughtfully. Productive and intentional critics like lawyers already exist to partner with government innovators to appropriately explore and suggest ways to mitigate against harm. In my government career, great lawyers have been key to any truly innovative progress I have made, and often have become some of my best friends as we “solve together.” Most importantly, when adding new “critic classes,” we need to be very careful about what the incentive structure of that critic class is. If they hold, maintain, or grow their power by driving fear, complexity, and uncertainty, then don’t be surprised when the “doers and innovators” flee that organization and progress grinds to a halt.

Countermeasure 9. Rules Require Reason: Smart Over Strict

Learning by doing in small, controlled, and intentional ways is one of the best methods to understand the risks and develop mitigations.

If possible, implement pilot programs to test new guidelines and technology. This approach allows for cautious experimentation while still enabling progress, reducing the fear of committing to large-scale changes. These experiments also provide evidence that can be critical in shaping options, priorities, and approaches and demystifying assumptions. By putting out practical and safe AI tools, we are able to see how people use them, confirm that they use them responsibly, and simplify the guidance and develop more effective training and education. It is much more efficient to see what people do on a small scale and adapt and iterate than it is to assume what they will do on a large scale and over-engineer guidance or training for behaviors or gaps you haven’t actually verified. “Verify before vilifying.”

This is why in AI education, we think about the four stages of a relationship with technology: “Ta-Da,” “Uh-Oh,” “Ah-Ha,” and “Ho-Hum.” The “Ta-Da” is when someone gets to use, access, or try the technology and see what it does. The “Uh-Oh” is when, after trying it, the practical experience helps the person think through how the technology might impact them, their work, or their team, and assign questions, intentions, and thoughtfulness around its application. “Ah-Ha” is when they find the intimacy of the value in their life and realize where time might be saved, hours given back, and informational advantage gained. Finally, “Ho-Hum” is where the technology becomes assumed as part of our lives, to include identity and practices like “Googling it.”

Taking people through this journey with intention is how permanent technology adoption occurs. Without trying, there can’t be great questions; without great questions, there can’t be great answers; and without great answers, there can’t be great change.

Tactic 10. Question Propriety: The Jurisdictional Quagmire

Saboteurs were encouraged to raise concerns about whether certain actions were within a particular group’s jurisdiction, or if they might conflict with higher policies or external opinions.16

In modern practice, questioning the propriety of actions, raising concerns about jurisdiction and compliance with regulations, and being overly cautious of potential conflicts with other entities can create significant roadblocks. An initiative to scale access to AI may be stalled because a committee member implies concern that the plan conflicts with other regulations or laws with no actual proof or familiarity with the regulation or law. The resulting review often takes months, and the initiative or technology adoption loses momentum and attention, leading to inaction on an important opportunity.

Countermeasure 10. Own the Turf: Cut the Power Struggles

Create a clear arbiter to resolve questions of propriety or jurisdiction.

Clearly define the roles and responsibilities of each team or committee at the outset of any project. An initiative or project charter can be helpful to clarify who is expected to do what and who has the power to make decisions.

Encourage collaboration and communication between departments early in the planning stages of any project. By involving relevant parties from the beginning, potential jurisdictional conflicts can be identified and resolved before they become roadblocks. A great way to do this is to introduce a “stakeholder press release” in which all stakeholders describe what they anticipate they will be proud of when the initiative rolls out, what it will achieve, and the impact it will have. This can often help surface equities and get teams on the same page in terms of why what they are doing matters.

Encourage the same stakeholder group to write a short “pre-mortem” outlining the reasons an initiative would end. Encouraging stakeholders to think about when and why they would stop doing a process, procedure, initiative, or other element makes it easier for everyone to know when it has run its course.

We often reward starting new things in government, but the true heroes are those who are brave enough to stop things that have become engrained and acceptable sabotage.

Tactic 11. Delay Action: The Slowdown Strategy

The manual advised delaying the delivery of orders, to ensure nothing was completed in a timely manner.17

Whether due to “red tape,” lack of urgency, or intentional stalling, delays in government operations are notorious. A federal contract may be approved, but the final paperwork is held up in administrative processing for weeks, causing costly project delays. How often have we heard “This sat on their desk for weeks” or experienced production bottlenecks when a worker went on vacation?

Countermeasure 11. Speed Over Stall: Have a Bias to Action

Break down large, multi-phase projects into smaller, actionable milestones with clear deadlines. Use project management tools that transparently track task completion, progress, and product delivery. These tools can provide real-time visibility into delays and help identify where action is stalled.

Recognize the difference between activity and progress: having a meeting is an activity, deploying a new requested feature is progress. Ensure your teams understand the difference between the two.

Create a culture that celebrates people and leaders who realize when it might be time to call it quits on an initiative and move on.

In all parts of a project, always have an eject button. Sunk-cost fallacy, a cognitive bias in which individuals continue to invest in a project, decision, or activity not because it is likely to succeed but because they have already invested significant time and resources, is strong in large organizations. Instead of basing the decision on future real benefit, they reflect on the past costs “sunk” where they cannot be recovered.

In today’s speed-chess world, we can’t afford to keep trying to move a program down the line when it has been overtaken by events or better technology, when a leadership transition removed the executive champion, or other issues. “Slow death,” when a project, process, or initiative is slowly starved of time, resources, and attention, is much harder on an organization, its people, and its culture. Create a culture that celebrates people and leaders who realize when it might be time to call it quits on an initiative and move on.

Don’t expect “one-size, fits all.” Organizations crave “enterprise solutions,” but often those solutions work for one part of an organization or a group at a particular maturity level, but not for everyone. The right tool for the right challenge at the right time and right value is critical to continuing to get the best technology into the right hands. Some people will need advanced options; others, just the basics.

Tactic 12. Assign Unimportant Jobs First: Misaligned Priorities

The manual suggested assigning unimportant jobs first and giving critical tasks to inefficient workers with outdated equipment.18

Important tasks are often delayed in favor of smaller, less critical work. This misalignment of priorities can lead to inefficiency and missed opportunities. A government information technology department, for example, may spend weeks on minor software updates while a critical cyber security overhaul is delayed. Often this is due to a lack of strategic time. If there isn’t any bandwidth to design good interventions, then neither the confidence nor the plan is in place to execute it. With the constant pressure of long lists of problems mounting up, it is easy to focus on the ones that take the least effort to act upon, while leaving other big initiatives unaddressed or outsourced to third parties. Even outsourcing becomes problematic because a customer doesn’t have the time to correctly identify and put a framework around what success is, leading to an implementation partner or vendor set up to fail, and predictably, budget and time overruns on the project.

Countermeasure 12. Real Work First, Busy Work Later

Great cultures require a clear “north star” and a way to orient our actions around it. Asking myself “Would I Do This in War?” is a simple way of helping me to decide if something truly matters. Create a valid and simple way for you and your team to know and identify what advances your north star.

Implement a task-prioritization system that ranks tasks based on importance and urgency. Critical tasks should take precedence, and resources should be allocated accordingly.

Doing the “impact math” or “toil calculations” can help you prioritize things that make a difference. If you are not sure what those things are, break down the metrics around the time, cost, hours, resource savings, the number of people impacted, or the impact on mission. Impact math is critical to ensure that everyone understands how the prioritization calculus works.

Impact math is also critical to disrupt “it’s just” language. Adding an element to the “Authority to Operate” process might imply that “It’s just one more tick box.” Does adding that one tick box provide enough cyber security risk reduction at the cost of getting new technology in the fight by ten days? That tick box on a stand-alone system with non-critical data may be an unnecessary barrier; on a critical system, it might be the difference to keeping our most precious knowledge from the adversary. Impact math helps us know the difference and make good choices.

One of the big differences I noticed while working in commercial environs is that commercial organizations understand the value of time, the cost of delay in getting something to market, the importance of getting a new feature added to their bottom line, and the need to secure their place in the competitive landscape. We should do the same. We need to consider both the benefits and the costs of our actions. We also must understand and be realistic where “toil” is being continued as a misguided form of job security either on behalf of employees, or by employees directly. There is enough work to go around, and not enough people to do it. Let’s focus on having people do the right work, not any work.

When we make people “figure it out,” we both set them up for failure and, more importantly, show our lack of respect for their time and their expertise.

Information technology doesn’t matter if it isn’t used. All too often, technologists fall in love with the technology rather than falling in love with the customers that technology is intended to help. At Google, we would say: “If a user mis-spells something and can’t find what they searched for, it’s our problem.” We didn’t blame the user — we understood that empathy and acceptance of human strengths and weaknesses were critical to putting out truly useful technology. Google also famously “dog-foods” their technology, meaning they use it before they put it out publicly; empathy grows quickly when it’s you being subjected to the process, training, or technology. Have “toil tours” where you, the leader, go and experience what your customers or people experience. I have never done a “toil tour” that hasn’t specifically provided me clarity on what we could make better as soon as possible.

Additionally, in government, we always underestimate the impact of the complexity/bureaucracy required to access a technology counter-weighted to its ability to make a difference. For example, if you have taken the time and resources to create a large and dynamic data set, but it takes people four forms and four weeks to get access, your data is unlikely to have the impact you wanted.

Similarly, in cyber security, as authorizing officials we often lament that people do not take the steps to appropriately manage their technology against cyber threats after we award the authority to operate. We assume that this is because “Cyber Security is not important to people.” What I often find is that, as experts, what is obvious to us is toil for others. If we want people to act, we should remove any toil that impedes the behavior that we want. So, if we want programs to do penetration testing on a system to ensure its cyber resistance, telling them what providers are available and providing the service purchasing mechanism makes it easier for them to meet the goal. When we make people “figure it out,” we both set them up for failure and, more importantly, show our lack of respect for their time and their expertise. Helping them do what is needed by making it easy for them is both empathetic and more likely to result in the outcomes important to our mission.

Tactic 13. Insist on Perfect Work: The Perfectionist Trap

Saboteurs were told to demand perfect work on relatively unimportant tasks, sending back items for minor refinishes — thereby wasting time and resources.19

Perfectionism, especially in low-priority tasks, is a common scourge of government that can lead to significant project delays. The pursuit of perfection in minor details often overshadows the broader objective. Early in my career, I remember a report on departmental efficiency I worked on being repeatedly delayed due to formatting changes and style manual edits, successfully delaying any scrutiny or accountability on that department.

Countermeasure 13. Progress Beats Perfect

Establish a clear standard of “good enough”; instead of demanding perfection up front, engage often and iteratively. This approach allows for continuous improvement without delaying project timelines.

Share with your team that you are happy to receive a “30 percent” document or concept, give rapid feedback, and allow them to further iterate. Teams that do this often have higher levels of trust, better communication, and faster results and outcomes versus teams who believe or fear that what they present must be “perfect” before it is shared with leaders.

One of the interesting elements I observed as I used generative AI is that I became a better prompt engineer or “question asker” to the AI when I was intentional and thoughtful about how I communicated the assignment, the reason, the goals, the details, and the outcome I was trying to achieve. While doing this, I realized I was giving the AI better and more thoughtful directions than I was giving my own people. When I started to be more specific and intentional around my information framing and directions with my teams and then allowed them to quickly iterate, bring options, and adapt, not only did we make more progress together faster, but also they were happier.

Tactic 14. Multiply Paperwork: The Bureaucratic Burden

The Simple Sabotage Field Manual recommended multiplying procedures and clearances, creating duplicate files, and requiring numerous personnel to approve every decision.20

Government operations often require multiple approvals, redundant forms, and excessive documentation, all of which slow down processes and frustrate employees. For example, to approve a small budget transfer between projects, a financial department must complete multiple forms, and obtain signatures from several levels of management and data entries for different offices. The entire process takes weeks, even though the request is straightforward and negligible. This is particularly chafing when, at the end of the fiscal year, one team has excess funding and spends it on unimportant things rather than moving it to another team with a critical need or opportunity because the paperwork required to move it is just too difficult.

In fact, most AI use cases are straightforward and simple and the technology is readily available.

Another example is the paperwork that many organizations are putting in place to navigate AI use cases. Because many perceive AI as a “one size fits all” monolithic technology, they often design process, forms, and oversight with the assumption that all AI use cases are the scary “equity and safety” kind. In fact, most AI use cases are straightforward and simple and the technology is readily available. For example, using AI for the purpose of translating information in different languages to make it available to everyone (something that AI is good at) should have a lightweight approval process that reflects the advanced stage of the currently available AI services and the positive impact that providing clear access to more people achieves. On the flip side, for AI use on something that has more sensitive issues like hiring decisions or fraud flags, the process should be more intentional and more robust.

Treating both use cases (Translation–Low Risk vs. Fraud Flagging–Medium Risk) with equal and overwhelming bureaucracy simply discourages the use of AI in safe, quick-win areas and results in less use of the technology where it can help the most.

Countermeasure 14. Lighten the Load: Trim the Paper Weight

Regularly review processes to identify redundancies, outdated requirements, or steps that can be automated or digitized. Simplify or consolidate forms and approvals wherever possible to reduce administrative burdens. Most government organizations already have an identity and authentication process built into their computer access systems. These digital workflows automate and streamline paperwork, so the same information does not have to be filled out repeatedly.

Similarly, cut bureaucratic hurdles by granting employees decision-making authority and reduce unnecessary paperwork by limiting the number of approvals required.

Employees should be incentivized to identify things their organization can stop doing. My favorite example is a team who implemented a “stop stupid” process where people could identify things that were getting in the way of what mattered.

Some organizations create “minus two rules” that require two steps to be taken out of a process every year. Others offer an incentive or “kill bonus” for employees who bring forward suggestions pertaining to things that the organization can stop doing or processes they can “kill” to save time and resources.

Recognize that it is heroic in government to suggest stopping something. Make heroes of those who are willing to ask: “Why do we still do this? Is this important?” We don’t have to demonize past decisions, systems, or processes. They came into existence for a reason, but it is also okay to say those reasons may no longer be relevant or appropriate now.

Encourage people to use AI where the task requires someone to “make more words out of a few” or “fewer words from more.” AI is particularly good at this.

Tactic 15. Give Incomprehensible Explanations: The Fog of Bureaucracy

Finally, the manual suggested that individuals provide lengthy and incomprehensible explanations.21

In documents, communications, or reports, complex, highly technical and unclear explanations often obscure issues rather than clarifying them, leading to confusion and inaction. Policy updates are communicated via lengthy and jargon-filled memos, leading to potential misinterpretation and errors in implementation and even more inaction.

Especially in cases of new or innovative technologies like AI, the policies are so numerous and complex, and the guides so lengthy and cumbersome, that they dissuade even the most ardent AI fans from acting. Even AI experts’ eyes will glaze over after about two pages into most policy documents that are full of words that don’t say anything specific about what a practitioner can or should practically do.

Instead of looking at the elements of AI that are truly unique, different, or additive, and making specifically relevant alterations, many infer that AI requires entirely new data protection, privacy, security, authorization, or cyber processes, even though these processes are often some of the most comprehensive and complete in government and simply need small additions to meet the mark.

Countermeasure 15. Make It Clear, Not Confusing

Implement a plain language policy that requires all communications, especially those involving policies and procedures, to be written clearly and concisely. Avoid potentially confusing jargon and complex language. Highlight the additive areas where small, limited adaptations are needed to reflect technical advancement versus suggesting that entire processes must be designed from scratch. AI is great at making complex language more simple and straightforward.

Provide executive summaries highlighting the key points in simple terms for lengthy or complex documents. Summaries can convey main messages without requiring readers to sift through dense text.

Lastly, use visual aids. Incorporate diagrams, flowcharts, and other visual methods into explanations to translate complex information more simply. Visual representations can often convey ideas more clearly than text alone, reducing the likelihood of misunderstandings.

Stopping Self-Sabotage

Is the government’s greatest saboteur itself?

The sheer number of parallels between the Simple Sabotage Field Manual and the current government standard operating procedure illustrates the potential impact of these seemingly small daily frustrations. The government has no shortage of spies, saboteurs, adversaries, and unconventional outside threats, but no single external threat could cause as much wide-scale inefficiency as we do in our own organizations.

If time is a weapons platform, we have abandoned it, left it to rot and rust, and harvested it for parts — but it’s not too late to change this.

Fewer meetings, fewer emails, less bureaucracy, less policy — less of everything just might bring us more of everything. More security, more creativity, more innovation, more purpose, and more national security advantage.

Undoing these excessively complicated regulations and procedures will require full-scale reform, but it may be less complicated than we imagine. Each of us can start with ourselves. We can choose to be toil makers or toil takers, saboteurs or saviors.

Our people need relief, they need purpose, and they need passion and enthusiasm for our critical mission. Until we tackle the toil and refuse to let ourselves be saboteurs, we cannot hope to have the incredible national security advantage that is locked up in the precious minutes and passion of our people.

We designed this mess — we can get ourselves out of it.

Alexis Bonnell is the chief information officer and director of the Digital Capabilities Directorate at the Air Force Research Laboratory and a proud public servant. The views expressed in this essay are those of the author and not those of the Air Force Research Laboratory, the U.S. Air Force, the Department of Defense, or any part of the U.S. government, but she’s doing her best to inspire others to choose solutions, not sabotage.

Image: Adobe Stock, licensed