Paper #43. AI-assisted Confrontational-Physical Strategies: Human-directed Savant X Seeker

  • Thomas A. Drohan, Ph.D., Brig Gen USAF ret.
  • Commercial, Cyber, Leadership, Security, Strategy
  • No Comments

Continuing our march through the eight basic combinations of strategy introduced in Paper #39 (The Strategy Cuboid), we focus on confrontational-physical competitions (preventive and causative). We‘ll use Savant X Seeker’s hyper-dimensional relationship analysis as a research assistant. 

The text corpus continues to expand as I add more curated reports and articles. The sample, however, is limited to English-language or translated works that are publicly available and easily downloadable. The latter criterion currently eliminates China-based think-thanks. I’m looking for one that’s not a Party mouthpiece.

This collection includes ACCORD, African Centre for the Study and Research on Terrorism, Aljazeera Centre for Studies, Asan Institute for Policy Studies, Atlantic Council, Australian Strategic Policy Institute, Canadian Global Affairs Institute, Carnegie Endowment for International Peace, Centre for Public Policy Research, Center for Strategic and International Studies, Chatham House, European Institute of the Mediterranean, Hague Center for Strategic Studies, Japan Institute of International Affairs,  Korea Economic Institute, National Security Journal, Nakasone Policy Institute, RAND Corporation, Rajaratnam School of International Studies, Singapore Institute for International Affairs, and the US Institute of Peace. 

The initial search used the same words to represent the broad dimensions of the Strategy Cuboid: cooperate, confront, physical, psychological, prevent, produce. 


Viewed in the SmartNav tab, this search also produced 1000 results of highlighted words, sortable by Phrases, Terms and Concepts in the Smart Filter


Figure 1: SmartNav Text Search & Filters

From the list of Phrases, I selected Weapons Systems. This reduced the 1000 results to 27. This word might be too narrow a type of confrontation-physical competition, but with this in mind, it’s a viable start. 

Figure 1: SmartNav Text Search & Filters

Whereas the results of our previous search for cooperative-physical supply chains centered on securing trust and managing risk, this search for confrontational-physical weapons systems yielded topics related to robotics and autonomous systems (RAS). There were nine issue areas:

  1. Controlling RAS
  2. Corruptibility of data to attack
  3. Impact on materials usage rates
  4. Impact on lowering the red-line of violence
  5. Reducing the transparency of decision making, especially NC3
  6. Continuity of lower quantity higher-tech weapons systems investments 
  7. Responsiveness of C2 against high-velocity weapons
  8. Unsurmountable advantages for defectors from norms undermine stakeholder cooperation
  9. How to maintain human oversight in technologically interconnected rapidly lethal circumstances even if AI is used directly for non-lethal applications. 

I wanted to expand consideration of confrontational-physical competition beyond lethal weaponry because competitors fill that space to circumvent military superiority. While the latter is necessary to win kinetic battles, it’s insufficient to create superior combinations of effects. So I looked for appropriate Smart Filter Phrases and found the best one to be national security. That produced 52 results. The following content slightly broadened the weaponry aperture beyond the institutionalized perspective of traditional armaments: 

  • Protecting data and networks
  • Going on offense to develop emerging technologies such as new energy vehicles in order to set global standards in transportation and supply chains
  • Increasing participation in regional economic organizations and development so as to confront Chinese influence
  • Consideration of a unilateral US ban on assassinations (presumably of non-combatants, not killing a clear combatant such as Soleimani) and on interference in democratic processes (as distinct from authoritarian processes)
  • Countering hostile actions via and against new technologies in cyberspace, space, and the information environment (disinformation)
  • Preventing armed conflict by deterring, de-escalating and denying territorial invasion
  • Restricting AI partnerships to non-predatory (read, non-China) participants, preventing regional domination by any power (read, non-democratic) with partnerships that supposedly share that common goal (often untenable)
  • China’s civil-military fusion of technology constitutes a “war economy” 

The latter issue of civ-mil fusion contrasts with the West’s idealistic segmentation of competition into two spheres. First, there is the red line between peace and war as defined by the use of violence rather than by effect on the population (as in cyber-attacks). Second, government intervention in the economy is deemed to be unfair according to liberal economic principles that every nation-state selectively ignores anyway (tariffs, subsidies, etc.). Government intervention for “security” (the absence of threats to acquired values) is not regarded as unfair competition.    


Next, I de-selected national security as a Phrase and searched the Terms in the Smart Filter. Of those, I sequentially selected securityforces, and threats to find examples of confrontational-physical strategies. This generated 314, then 97 then 23 results. 

The focus of these passages was military threats, violence against social leaders, state-sponsored subversion, and externally stoked domestic violence. These all reflected the West’s “fight when deterrence fails” military approach to security even as China builds up its economic and financial foundations of military power. Confronting threats within the rule of law is almost always portrayed as cooperation in the West. In China, it’s rule by law. 

The West’s “fight when deterrence fails” idealism deflects operations away from proactive targeting because the effects must meet a higher standard than the law of armed conflict: evidence in a court of law. Authoritarian political systems have the advantage with respect to initiating threats to disrupt competitors compared to democracies that deter and respond to threats. The issue of radicalization and online terrorism stretches the legal systems of democracies, although the tension is leading to more intelligence-sharing among government and international agencies. 

Meanwhile, attacks carried out by criminals and terrorist groups in weak states, such as the abduction of children by Boko Haram Nigeria, Niger, Chad, and Mali, reinforce the convergence of political governance, military professionalism, economic development, and social violence. Without a competitive economic, financial and social foundation, military superiority is fleeting and irrelevant.  


Finally, I retained forces and threats from the Terms search above and searched the Concepts (paired terms) in the Smart Filter which had generated as its top result, development…technology. I selected forces…threat. This produced a manageable 30 results. Investigating these, I found that all of them were characterized as military and para-military. Political, economic and social forces and threats were absent unless tied to violence.  

In the RESULTS passages, conventional concepts of deterrence were expected to work against conventional and missile threats, but not against cyber attacks, terrorism to incite extremism, and disinformation. China’s conventional or cyber-attacks (via backdoors) against technology manufacturing sites (microchip plants in Taiwan and South Korea) were cited as potential concerns, perhaps enough to form a “techno-democracy” alliance. 

This thinking, however, ignores how economics can drive alliance alignment at least as much as political values. Clearly, what was missing was the consideration of the same tools that authoritarian regimes use against democracies—proxy or constructed (“islands”) territorial invasion, manipulated markets, digital theft, official disinformation, and so forth.

The mixing of nuclear deterrent concepts with non-nuclear forces is risky. The possibility that China deliberately seeks “warhead ambiguity” between nuclear and non-nuclear weapons as a strategy to deter US attacks on its conventional forces increases nuclear instability and risks provoking an attack under uncertain circumstances. Add to that, increased precision of weaponry, cyber attack anonymity, and unknown reliance on RAS by an authoritarian state paranoid about domestic upheaval.    

The next step in the analysis was to look at linkages as presented visually in HyperNav.


In the HyperNav tab, the original search revealed the following relationships:

Figure 2: HyperNav First-order Relationships (Low Node Control Setting)

Instead of the United States at the center of the linkages as in Paper #42, development emerged. At first, I thought that this was due to the larger international corpus. However, my earlier search word choices included “cause” whereas this search replaced that with “produce.” Trying out similar but nuanced search terms and comparing the results shows the disproportionate effect of apparently small changes in inputs. In this case, it’s an ecosystem of words.   

Here are the overall shapes of the nodal searches set at minimum, mid-range, and maximum Node Control settings: 

Figure 3: HyperNav First-Order Relationships at Minimum Node Control Setting
Figure 4: HyperNav First-Order Relationships at Mid-Node Control Setting


Figure 5: HyperNav First-Order Relationships at Max Control Node Setting 

For the next search, I gradually increased the node control to look for confrontational-physical indicators. Set at max nodes, I selected forcethreats and conflict. Separately this produced 231, 125, and 88 passages, respectively. I decided that 231 is too many to go through for this human right now. However, here’s a sampling of the threats found by Seeker:

  • Anti-Microbial Resistance
  • Social effects of long-term US military intervention in Iraq and exploitation by Iran, dependence on foreign supplies of munitions
  • Cyber attacks against system integrators of data and defense manufacturers
  • Use of surveillance and tracking via devices in an Internet of Behaviors arena
  • Hybrid warfare
  • Election interference
  • China’s capability gaps in informatized reconnaissance-strike wrt competitors with territorial claims
  • Increased missile accuracy and range
  • State violence against human rights defenders and social leaders
  • External intervention therefore the need for a nuclear deterrent (North Korea)
  • Cyber offense with nebulous ties to state military forces
  • Turkish military action in Libya to support the regime and illegal drilling near Cyprus can isolate Turkey from NATO
  • Violent extremism’s causes—transnational crime, illegal migration, and radicalization which is difficult to manage consensus on in democracies
  • China’s integration of defense, economic and public diplomacy agendas with an understanding with Russia,
  • AI connectivity, autonomy, faulty data, opaque algorithms, and software updates
  • Complex online platform ecosystems that complicate effective law enforcement; therefore the EU’s Digital Services Act and Digital Market Act seeks to regulate the size and access to data that allows one competitor to dominate
  • China’s interference with US unmanned underwater vehicles or uncrewed maritime systems (both misnomers) due to SSBN vulnerability to unmanned UUV USV could prompt China to attack it  

To find actionable relationships connected to such diverse threats, I selected force and threats, which produced 40 passages. In general, democracies refuse to counter new threat technologies in kind, except for military capabilities against military threats. This is a gaping vulnerability. Indeed conflicts that involve the use of force against threats include, but extend well beyond military forces:

  •  How to defend satellite (military and commercial) networks
  • How to improve governance in weak states where climate change erodes the capability to govern and this is exploited by terrorists and criminals
  • China’s nuclear force modernization and authoritarian, coercive diplomacy intended to split the US from allies with respect to the military, political and economic defense of Taiwan
  • China’s increasing of the “zone of contest” in the Asia maritime and air commons
  • Russia’s physical imposition of influence within former USSR states relies on controlling escalation to increase and decrease various target audiences’ will and capabilities to fight; this logic is being applied to social and economic relationships as well
  • Democratic states’ future force planning to meet operational objectives and fill mission capabilities gaps are politically constrained in closing the larger gap—the more-than-military consequences of such objectives and missions

Combined Effects, Not Just Combined Arms

The reluctance of democracies to combine DIMES (Diplomatic, Informational, Military, Economic and, especially, Social) effects such as coercion and persuasion is based on the assumption that the latter could negotiate away the advantages of the former. A combined effects perspective—all-effects and all-domain, rather than a combined arms perspective—all-domain, solves this problem. By specifying the objective (de-nuclearizing North Korea, for instance), coercive and persuasive effects can produce a dilemma for Pyongyang rather than undermine each other. Democracies have issues sustaining dilemma-inducing strategies over time. This shows up repeatedly in the passages of the corpus when we look for relationships that produce effective strategies. 

For instance, with force and threats still selected, INSIGHTS generated the following nodes: supply chain, raw materials, and RAS. Consider some of the relationships across the dimensions of strategy.

Dimensions of Strategy, Not Just ”Offense” and ”Defense”

An example is coop-frontational Chinese supply chains that weaponize raw materials, manufacturing and surveilled demand. In response, democracies consider the use of “defensive” and “offensive” measures. INSIGHT-discovered passages included “defensive” lawfare that generates anti-dumping and countervailing duties and “offensive” strengthening of R&D, education, and manufacturing. Public opinion tends to see the former as confrontational but not the latter. Warfare, after all, is defined in terms of violence. 

The language of defense and offense is not very useful compared to the distinctions of: cooperative—confrontational & physical—psychological ends, ways and means; and preventive—causative effects.

If we take a confrontational-physical approach to prevent and cause, for instance, anti-dumping and countervailing duties, R&D, education and manufacturing become weapons to prevent and produce strategic comparative advantages.

How are they confrontational? By targeting a competitor’s vulnerabilities. Such as a persistent information campaign to: (a) expose predatory activities (China’s economic coercion against the Philippines and Australia, e.g.); and (b) exploit the absence of allies (more engagement, not less, than authoritarians can manage).

How are they physical? United Nations institutions built to promote and defend human rights, and infrastructure for sustainable development are the arenas of competition. They confront systemic vulnerabilities and global threats with international standards and participation. This is a very different perspective than the prevailing “deter then fight” approach to warfare.    

Next, I selected threats and conflict which produced 19 passages, mostly about violent conflict with some mention of hybrid combinations. Related to threats and conflict are: China’s use of development banks to shift power from West to East; China’s use of international institutions to reshape global governance without human rights; Russia’s disinformation and geopolitical influence; and Turkey’s re-engineering of identity in Syria. The roots of confrontation are poor governance, economic failure, ethnic or religious intolerance, territorial disputes, and radicalization.  Preventing escalatory regional conflict is a top priority, with complex confrontations simply regarded as humanitarian crises. 

Selecting INSIGHTS here took me to companies, businesses, private sector and raw materials, in that order based on node size. These were suggestive, without pointing toward passages in the corpus. So I selected force and conflict. This led to three interesting insights.

The first was a concern about the use of partners, not proxies. Working with a partner implies closer command and control relationships than the Russian tactics of hands-off chaos. The higher standard of partnerships, however, leads to the development of and entablement with civilian competencies. Depending on the state of civ-mil relations, that often leaves non-civilian (police, military) security to local tyrants, proxies or foreign forces. Remember how it took for South Koreans to institutionalize electing a non-military President?

The second insight was the relationships among Russia, companies, and firms that assisted in China’s development of a second-strike nuclear capability. The historical context was the Sino-Soviet split, disastrous Cultural Revolution in China, post-Mao recovery, and the emergence of ideological control based on economic growth. China’s overtaking of Russia (except in numbers of nuclear weapons) poses the question, what will deter a self-sufficient nuclear China from territorial seizures? 

The third point concerned cyber force competition to acquire zero-day vulnerabilities on competitors, avoid attribution, and disguise attacks as normal procedures. Again, the prevalence among Western policymakers and academics of saving such capabilities to deter or prevail in conflict contrasts with authoritarian practices of effectively waging cyberwar all of the time.


From the analyst to the policymaker, a key question about competition in complex environments is, how do we filter what is and what isn’t a real threat? Use of force is the legacy answer. Unfortunately, that misses most threats. Most threats are not uniformed military attacks and include confrontational physics that produce violence indirectly. Lethal effects are nth-order such as deaths due to: attenuated logistics in military operations; sabotaged energy access in urban areas; reduced health care and sanitation; disrupted agriculture; and aggravated social conflicts. 

With the advent of Joint All-Domain Command and Control JADC2, the US needs an expansion of security strategy to stay competitive. I’ve argued for the Strategy Cuboid.

Unless commanders are prepared and empowered to influence relationships in the information environment besides blowing stuff up, we can have an exquisite C2 system and still end up losing complex wars.  

This is understandable but not insurmountable. Why?

Other than confrontational-physical strategies to reduce the risk of conflict, such strategies to achieve objectives in non-war security competition are hard to sustain in democracies. One bold approach is the UK’s three-pronged strategy, Global Britain in a Competitive Age. As it relates to data to reduce risk, the effort includes developing technical standards, regulating infrastructure and commercial activities, and establishing global rules of behavior. European efforts in the Middle East promote economic development and reduce actions that exacerbate conflict, subject to divisiveness over competing interests (in Libya, France supports Khalifa Haftar and Italy supports the UN-recognized government).

In contrast, authoritarian states are able to use confrontational-physical strategies in more circumstances—proxy and direct invasion of territory, energy and trade cutoffs, cyber theft and disruption, assassinations of non-combatants. 

Moreover, confrontational-physical strategies claim cooperation, particularly with respect to targeted audiences that matter to their power base. The latter usually matter domestically to whip up nationalism and abroad to create divisiveness. The results are: invasions and ethnic cleansing protect Russian and Chinese identity; energy and trade cutoffs to incentivize agreement with prices or policies; cyber theft and disruption shamelessly deflected by deceitful promises of enforcement; assassinations flatly denied and follow-on sanctions decried as unjust.

The use of cooperative-confrontational-psychological strategies complement these claims and are the subject of the next Paper in this series.   

Author: Thomas A. Drohan, Ph.D., Brig Gen USAF ret.

Leave a Reply