Article published in the Sep 26 edition of “News Analytics” magazine
The character of warfare evolves alongside technological innovation. Today, another transformative revolution is reshaping global security. The integration of artificial intelligence (AI), autonomous weapons systems, and human-machine teaming into military operations is underway.
Deterrence has traditionally depended on a human adversary believing that the costs of aggression will outweigh its expected benefits. Nuclear weapons made this logic especially powerful by creating the prospect of unacceptable retaliation.
Autonomous warfare, however, is altering the speed, visibility and psychology of military competition. Artificial intelligence-enabled systems can detect, classify, track and engage targets with limited human intervention. In contrast, autonomous drones, loitering munitions, cyber tools, underwater vehicles and defensive systems may operate across several domains simultaneously. The central question is whether autonomous systems will replace conventional or nuclear deterrence or alter its dynamics.
The Transformative Effects on Deterrence
At the core of this transformation is the integration of AI across the full spectrum of military functions, including intelligence, surveillance, and reconnaissance (ISR), targeting, command and control, logistics, and, increasingly, the application of force. Autonomous platforms (ranging from loitering munitions and drone swarms to AI-enabled decision-support systems) operate with varying degrees of independence once activated. Human-machine teaming seeks to combine the strengths of both.
Autonomous systems are particularly suited to denial because they can support persistent sensing, distributed defence, and rapid disruption. Unmanned aerial vehicles can monitor approaches to military installations; autonomous underwater systems can observe maritime activity; ground robots can support border surveillance; and AI-enabled command systems can fuse information from multiple sensors. A networked force may complicate an adversary’s effort to achieve surprise, suppress defences or destroy high-value targets.
Autonomy can enhance capability by allowing military systems to process enormous quantities of data and act faster than human operators. An AI-enabled surveillance network may identify changes in an adversary’s deployment patterns before traditional intelligence systems do. Autonomous platforms can then maintain continuous patrols, coordinate with one another and respond to selected threats without waiting for detailed instructions. Such capabilities may strengthen deterrence by denial.
The importance of denial will increase as military forces become more dispersed. Instead of protecting a small number of vulnerable platforms, states may deploy large numbers of mobile, concealed and networked systems. An adversary would then face a difficult targeting problem: destroying some autonomous platforms would not necessarily turn off the entire force. This resilience can reduce the attractiveness of a first strike.
Autonomy can influence credibility. A state with resilient, dispersed, and relatively inexpensive autonomous systems may be able to respond to aggression even after suffering damage to its command centres, air bases, or naval facilities. Swarms and unmanned systems can generate operational effects without exposing large numbers of personnel to danger. This may make retaliation more politically acceptable and therefore more credible.
Autonomy can also impose operational costs. Defensive systems may use algorithms to detect incoming missiles, drones or aircraft and recommend or initiate responses. If these systems are reliable and their employment conditions are clearly defined, they can reduce the prospect that an adversary will achieve a quick victory. The strategic message becomes: aggression will encounter a persistent, adaptive and difficult-to-suppress defence.
Escalation Dynamics
The most serious challenge is compressed decision time. Autonomous systems can identify and respond to threats faster than human institutions can verify information, consult political leaders or establish whether an incident was deliberate. In a crisis, this may generate a “use-or-lose” mentality. Commanders may fear that delaying action will allow an adversary’s autonomous systems to destroy their own sensors, communications or retaliatory forces.
Machine-speed operations can therefore create pressure for pre-delegation. Political and military leaders may authorise automated responses in advance because human approval would be too slow. Yet pre-delegation carries significant risks. An algorithm may misinterpret a civilian aircraft, a training exercise or a cyber intrusion as an attack. A technical malfunction could trigger a chain of responses that neither side intended.
Autonomous warfare also creates the possibility of machine-to-machine interaction. One state’s defensive algorithm may classify the activation of another state’s autonomous system as hostile. The second state may then respond automatically, producing reciprocal escalation. Unlike human decision-makers, algorithms do not possess political intuition, historical memory or an inherent preference for restraint. They optimise according to programmed objectives and available data. If the data are incomplete or manipulated, the resulting decision may be technically rational but strategically disastrous.
Strategic Stability
Autonomous warfare will have its most consequential impact on strategic stability when it interacts with nuclear forces. AI-enabled systems may improve early warning, intelligence analysis and the protection of nuclear assets. They could help identify suspicious activity and strengthen command-and-control resilience. In this respect, autonomy may reduce uncertainty and support more informed decisions.
The opposite is also possible. AI-enabled surveillance and autonomous strike systems could threaten mobile missiles, submarines, command centres and communications networks. If one state believes that its nuclear deterrent is becoming vulnerable, it may adopt more aggressive readiness postures or delegate greater authority to military commanders. The combination of improved detection, persistent tracking and rapid attack could generate fears of a disarming first strike.
This danger is not limited to actual technical capability. Perception also does matter. A state may respond to what it believes an adversary can do, even if the adversary’s systems are not as effective as assumed. Strategic competition could consequently become unstable through exaggerated assessments of AI-enabled counterforce capabilities. The combination of advanced intelligence, strike systems and missile defence may challenge the traditional assumption that retaliation remains assured after a first strike.
Autonomous systems may also blur the boundary between conventional and nuclear operations. An attack on dual-use command-and-control infrastructure by conventional autonomous weapons could be interpreted as preparation for a nuclear attack. If a state cannot distinguish between an anti-conventional operation and an anti-nuclear operation, it may escalate rapidly. Maintaining clear separation between nuclear and conventional assets, preserving human decision authority and establishing crisis communication channels will therefore become increasingly important.
Implications for India and Regional Security
For India, autonomous warfare is directly relevant to a security environment characterised by contested borders, maritime competition, terrorism, cyber threats, and the possibility of simultaneous pressure from multiple directions. Autonomous systems can enhance surveillance along difficult terrain, improve maritime domain awareness and support the protection of critical infrastructure. They can also strengthen deterrence by denial by making surprise incursions, drone attacks and limited aggression more difficult to execute successfully.
However, technology alone cannot provide deterrence. India would require an integrated architecture combining sensors, secure communications, electronic warfare, cyber resilience, air defence, space-based support and trained human operators. Autonomous platforms must remain connected to a wider command-and-control system that can function even when communications are degraded, or networks are attacked.
Concluding Thoughts
The transition to algorithm-assisted autonomous warfare will not change the basic concept of strategic deterrence. It will mandate a more controlled version of it. One that accounts for machine-speed decision cycles and distributed, ambiguous chains of responsibility. It will also have to cater for an adversary whose calculations are also based on software architectures. Autonomy will create an additional layer of deterrence, one based on persistent surveillance, rapid response, distributed force structures, denial of objectives and the ability to impose costs at machine speed.
As autonomous capabilities continue to grow, policymakers will have the important task of adapting and finding balance during this transition. In autonomous warfare, deterrence will depend not just on cutting-edge technologies but also on ensuring these systems operate within clear, transparent guidelines, robust command structures, well-rounded legal frameworks, and internationally accepted norms. The goal isn’t to slow technological progress, but to ensure that humans, who are ultimately responsible for warfare, still have the power to choose peace.
Please Add Value to the write-up with your views on the subject.
For regular updates, please register your email here:-
References and credits
To all the online sites and channels.
Pics Courtesy: Internet
Disclaimer:
Information and data included in the blog are for educational & non-commercial purposes only and have been carefully adapted, excerpted, or edited from reliable and accurate sources. All copyrighted material belongs to the respective owners and is provided only for wider dissemination.
References: –
- Institute for Defence Studies and Analyses. (2019). Artificial intelligence and national security: Indian perspectives. Manohar Parrikar Institute for Defence Studies and Analyses.
- Rajaraman, V. (2014). Robot soldiers: Artificial intelligence and the future of warfare. Resonance, 19(11), 1037–1048.
- Horowitz, M. C. (2016). The ethics & morality of robotic warfare: Assessing the debate over autonomous weapons. Political Science Quarterly, 131(2), 331–351.
- Boulanin, V., Saalman, L., Topychkanov, P., Su, F., & Peldán Carlsson, M. (2020). Artificial intelligence, strategic stability and nuclear risk. Stockholm International Peace Research Institute.
- Su, F., Wan, W., Saalman, L., & Chernavskikh, V. (2025). Pragmatic approaches to governance at the artificial intelligence–nuclear nexus. Stockholm International Peace Research Institute.
- Boulanin, V., Saalman, L., Topychkanov, P., Su, F., & Peldán Carlsson, M. (2020). Artificial intelligence, strategic stability and nuclear risk. Stockholm International Peace Research Institute.
- Johnson, J. (2020). Artificial intelligence, drone swarming and deterrence. Journal of Strategic Studies, 43(5), 1–24.
- Horowitz, M. C. (2019). When speed kills: Lethal autonomous weapons, deterrence, and stability. Journal of Strategic Studies, 42(5), 764–788.
