Business leaders keep asking the wrong question about artificial intelligence. They ask which model to buy, which chatbot to install, or which department can lose the most manual work. Those are procurement questions. The strategic question is different: How can an organization perceive change sooner, decide with better context, and move resources toward an opportunity before the market understands what changed?
That is where ideas from cognitive warfare and irregular warfare become useful, provided we translate them carefully and ethically. A customer is not an adversary. A market is not a battlefield. Manipulation, deception, and coercion have no legitimate place in a trusted commercial relationship. But the underlying disciplines of systems thinking, distributed sensing, asymmetric advantage, resilience, influence, tempo, and adaptation have enormous relevance to business.
The companies that understand this distinction will use AI to strengthen judgment and compress the distance between signal and action. The companies that do not will automate old processes, produce more noise, and wonder why the technology never becomes a durable advantage.
The contest is over perception, interpretation, and action
NATO's Joint Warfare Centre describes cognitive warfare as a contest involving human and machine cognition, with the power to shape how information is interpreted and how decisions are made. The commercial lesson is not to target the public with military techniques. It is to recognize that every organization already operates inside a cognitive environment.
Customers are deciding what to trust. Employees are deciding which signal deserves attention. Search engines and answer systems are deciding which source to surface. Executives are deciding whether a movement in the data is noise, risk, or opportunity. AI agents are increasingly making bounded choices about what to retrieve, summarize, recommend, route, or execute.
In that environment, the scarce resource is not content. It is coherent understanding. A company with accurate data, a clear narrative, trusted evidence, and short feedback loops can outperform a much larger competitor that has more people, more platforms, and less shared context.
THE BUSINESS COGNITIVE LOOP
- 01SenseCollect meaningful signals from customers, operations, markets, systems, and people.
- 02InterpretConnect signals to context, history, economics, risk, and strategic intent.
- 03DecidePut the right decision in front of the right accountable person at the right time.
- 04Act and learnExecute within clear authority, measure the result, and improve the next cycle.
Irregular strategy favors leverage over mass
The United States Army's recent handbook on understanding irregular warfare emphasizes systems thinking and indirect, asymmetric activity designed to create dilemmas and increase costs. In business, the ethical translation is straightforward: do not meet a larger competitor only where its scale is strongest. Change the economics and the tempo of the contest.
A regional company cannot outspend a global platform. It can understand a customer niche more precisely, publish more authoritative evidence, design a more useful digital experience, and respond faster because its data and decisions are closer together. A focused professional services firm may not have a thousand-person delivery organization. It can encode its best methods, automate coordination, and give every experienced employee the leverage of an intelligent operating layer.
This is asymmetric advantage. It does not require a gimmick. It requires choosing the few capabilities that change the ratio between effort and outcome. In the AI era, those capabilities are usually proprietary context, connected systems, process knowledge, fast evaluation, human judgment, and the authority to act.
Automation should distribute capability, not centralize fragility
Most companies still implement automation as a chain of brittle handoffs. One system triggers another. A field changes. The workflow breaks quietly. Nobody knows who owns the result. Adding a language model to that chain makes the failure more articulate, not more intelligent.
Irregular organizations survive by distributing capability while maintaining intent. Businesses should build AI the same way. Leadership defines objectives, boundaries, risk tolerance, and measures of success. Teams and agents receive only the context and authority required for a specific mission. Important actions remain observable and reversible. Exceptions return to accountable people.
This is why I have argued that AI agents must become an operating model, not another destination application. An agent should not be a universal employee with unlimited access. It should have a defined job, a constrained identity, approved tools, measurable outcomes, and a clear escalation path. The same principle underlies identity-first agent security: authority must be explicit before autonomy can expand.
NIST's AI Risk Management Framework organizes responsible adoption around governing, mapping, measuring, and managing risk. That sequence is useful because it forces the business to understand the system before it accelerates the system. Governance is not the brake on AI. Good governance is what allows useful autonomy to move faster without making the organization fragile.
Marketing becomes a sensing and response system
Traditional marketing was organized around campaigns. Teams planned, produced, launched, reported, and started again. That rhythm made sense when information moved slowly and media was expensive. It makes less sense when customer behavior changes daily, creative can be produced instantly, and AI can monitor thousands of signals continuously.
The modern marketing system should sense changes in search behavior, customer questions, sales conversations, service friction, competitor positioning, content performance, and revenue quality. It should connect those observations to business context. It should then recommend or execute the smallest useful response: improve a page, clarify an answer, adjust a journey, create evidence, route an opportunity, or stop spending money where attention is not turning into value.
This is also where cognitive discipline protects the brand. Generative AI makes it easy to flood every channel with plausible material. But volume without meaning weakens trust. The objective is not to occupy more attention. It is to earn attention by being useful, specific, consistent, and demonstrably credible.
That is why the conventional labor-heavy agency model is losing ground to connected AI marketing systems. Reporting after the fact is less valuable when a governed system can identify a change, assemble the evidence, recommend a response, and route it for approval while the opportunity still exists. Monstrous Media Group has made the same argument in explaining why systems beat campaigns.
The website becomes infrastructure for people and machines
For most of the web's history, a website was treated as a publication or a sales presentation. I have spent years treating it as something larger: the public edge of an organization's operating system.
That view changes the architecture. Content must be structured enough for search engines, answer systems, and agents to interpret. Identity and permissions must be designed into actions. Data from customer interactions must improve the next decision. Accessibility, speed, semantics, evidence, and security become operating requirements rather than launch checkboxes.
In AI Is Rewriting the Web and the Marketing System Around It, I described the emerging four-audience website: one that serves people, search, answer systems, and agents. That is not a speculative design trend. It is the logical result of machines becoming active participants in discovery and transactions.
A company that still builds a website as a collection of pages will eventually need to rebuild it as a collection of trustworthy capabilities. A company that starts with structured content, clean interfaces, reusable services, consent, observability, and controlled machine actions can adapt without replacing the foundation every two years.
Why I have been early to this shift
I started coding in 1993, when the commercial web barely existed, and learned early that the visible interface is only the surface of a much larger system. Networking taught me to think in protocols, dependencies, failure domains, and routes. Security taught me that identity and authority matter more than convenience. Marketing taught me that perception affects action. Defense and statecraft taught me to study systems, incentives, influence, resilience, and decisions under uncertainty.
Those disciplines were often sold as separate industries. I never saw them that way. A website is connected to infrastructure. Infrastructure is connected to identity and risk. Marketing is connected to data and behavior. AI is connected to all of it. When those layers are designed independently, organizations spend more and understand less. When they are designed as one system, every useful signal can improve the next action.
That is why the work can look a decade ahead of where mainstream web design, development, and marketing are going. It is not based on predicting a particular model or platform. It comes from designing around enduring conditions: information will continue to accelerate, interfaces will become more machine-readable, routine execution will become less expensive, trust will become more valuable, and human judgment will move toward the exceptions that matter.
What to expect next
First, websites will expose more structured actions to authorized agents. The best sites will remain beautiful for people while becoming easier for machines to understand and safely use.
Second, marketing departments will look more like operating centers. Humans will set intent, evaluate evidence, protect the brand, and make high-consequence decisions. AI will monitor, assemble, route, test, and execute bounded work across systems.
Third, organizations will reduce software sprawl. Leaders will stop buying isolated AI features and start building a governed intelligence layer across shared data and approved tools. The economic value will come from fewer handoffs, less duplicate work, faster response, and better use of human capital.
Fourth, provenance and trust will become competitive infrastructure. As synthetic content becomes abundant, customers and machines will reward sources that can establish authorship, expertise, freshness, evidence, and accountability.
Finally, strategy will become more continuous. Annual plans will not disappear, but they will be supplemented by systems that sense change and help teams respond without losing the larger intent. The organizations that learn fastest will not necessarily be the organizations with the most data. They will be the ones that turn selected data into shared understanding and controlled action.
FIVE PRINCIPLES FOR THE NEXT OPERATING MODEL
- 01Compete on clarityMake trusted context available before adding more automation.
- 02Design for leverageAutomate the coordination that prevents skilled people from doing high-value work.
- 03Distribute safelyGive teams and agents bounded authority with observable outcomes.
- 04Shorten the loopConnect sensing, interpretation, decisions, execution, and learning.
- 05Protect trustUse AI to make the organization more useful, not more manipulative.
Put power behind human capital
The point of AI is not to remove people from the business. It is to remove the friction that keeps capable people from applying judgment. A strategist should spend less time collecting screenshots. A designer should spend less time resizing the same idea. An operator should spend less time reconciling systems. A leader should spend less time waiting for a report that describes last month.
The strategic organization gives people better context, useful machine support, and the authority to act inside clear boundaries. That is how automation lowers cost without lowering standards. It turns human capital into a force multiplier instead of treating payroll as a problem to be minimized.
Continue through the connected system: Read how AI is changing the web and marketing, see why production AI is an economics problem, and explore how AI can improve marketing ROI.
Sources and further reading
- NATO Joint Warfare Centre: Cognitive Warfare
- United States Army: Understanding Irregular Warfare
- NIST: Artificial Intelligence Risk Management Framework
- NIST AI Resource Center: Govern, Map, Measure, and Manage
- World Economic Forum: AI adoption and workforce strategy
Turn signals into decisions, and decisions into controlled action.
If your website, marketing operation, data, and internal workflows still function as separate systems, start a conversation with Brad. We can map the operating environment, find the highest-leverage opportunities, architect the AI and automation layer, and put more power behind the people who already understand your business.