On July 6, speaking at the first Global Dialogue on Artificial Intelligence Governance in Geneva, UN Secretary-General António Guterres condemned so-called “killer robots” as “morally repugnant” and “politically unacceptable,” admonishing that the rapid convergence of artificial intelligence and military technology demands urgent global governance. Guterres is not speaking hypothetically: lethal autonomous weapons systems, or LAWS, are becoming customary tools of war. Powered by geospatial artificial intelligence (GeoAI), these systems can identify targets, interpret sensor and satellite data, navigate complex environments, and carry out military operations with increasingly limited human intervention. Guterres’s remarks highlight the ethical stakes of GeoAI, whose growing role in autonomous weapons has prompted debate about clear standards for accountability and international governance.

GeoAI in Practice

GeoAI refers to the fusion of artificial intelligence with geospatial data, and its origins are mostly civilian. The technology combines GIS (geographic information systems) and artificial intelligence in ways that more efficiently and effectively handle vast amounts of spatial data. GeoAI can be used for environmental monitoring or urban planning, but also for military purposes. Understanding how GeoAI functions helps explain why it has become such a powerful military tool. 

During warfare, GeoAI works through multi-sensor fusion, combining multiple data sources, satellite imagery, radar, lidar, GPS/GNSS positioning, and onboard sensors, to create a full picture of the combat environment. Convolutional neural networks (CNNs), a form of deep learning, can detect and classify military technology. Increasingly, this processing happens on the platform itself, known as edge AI, rather than on a remote server or command center, allowing systems to make decisions in real time even when communications are jammed or unavailable. Once the target is identified, navigation systems use digital terrain models (DTMs) to plot paths, while radar, lidar scans, and GPS/GNSS data guide the system across terrain to the target and execute the strike. The whole process can occur with little or minimal human involvement.

Unmanned Autonomous Vehicles (UAVs) can use Edge AI to process photographic and LiDAR data onboard, allowing them to make real-time adjustments.

As these capabilities have advanced, GeoAI has become a strategic imperative in the global competition for military advantage. The United States, the leader in the AI race, has invested heavily in GeoAI-based applications, including in the testing and development of LAWS. A collaboration between the Pentagon, National Geospatial Intelligence Agency, and private sector corporations yielded Project Maven, a program that develops AI tools for target reconnaissance and recognition in military theaters. While Maven has faced ethical criticism from the start, the U.S. military has remained committed to further integrating AI into its defense systems and building scalable AI infrastructure.

The U.S. argues that developing LAWS assisted by GeoAI is necessary to deter and respond to adversaries pursuing similar capabilities. Although LAWS remain difficult to define under international law, other major powers have already made significant investments in AI-enabled military technologies. China, for example, has reportedly developed an “Underwater Great Wall,” a network of AI-powered underwater sensors designed to enhance surveillance and submarine detection in the South China Sea. Meanwhile, the war in Ukraine has demonstrated the growing role of AI-enabled drones, autonomous navigation, and geospatial targeting technologies in modern combat, as both Ukraine and Russia have integrated increasingly sophisticated unmanned systems into military operations. These strategic developments encourage even greater investment in GeoAI, but they have also intensified debate over the role of autonomous systems in modern warfare. 

The Debate

Proponents of autonomous weapons argue that their use can actually reduce civilian casualties. GeoAI can process far more data than a human analyst, enabling militaries to carry out more precise strikes with less collateral damage. This was the case the U.S. made in a 2018 working paper to the UN’s Group of Governmental Experts on LAWS, arguing that limiting civilian harm was itself a reason to develop the technology further, not restrict it. The paper described several concrete examples including self-destructing mines and submunitions that render themselves inert after they’ve served their purpose, software that helps commanders estimate likely collateral damage before a strike is even launched, and precision-guided weapons that need far fewer strikes to hit a target. 

However, despite these advances, real-world deployments show that this improved precision does not eliminate the risk of catastrophic error. In the early days of the ongoing U.S.–Iran War, an attack on a school in Minab, Iran killed more than 150 civilians, the majority of whom were schoolgirls. U.S. officials have since indicated that American forces were responsible and that errors in GeoAI-assisted targeting contributed to the attack. Although incidental civilian damage is tragically common in warfare, the role of autonomous systems creates a unique ethical hurdle. In particular, it complicates questions of accountability: who bears responsibility when AI-assisted decisions result in catastrophic errors? Moreover, as military personnel increasingly rely on AI-generated recommendations, there is growing concern that time pressures and trust in automated systems may encourage what researchers describe as “cognitive off-loading,” reducing human involvement in decisions with grave human consequences.

Much of the discussion over the ethics of LAWS and GeoAI concerns “meaningful human control,” (MHC) a concept that is not well-defined but is often used to discriminate between various types of autonomous weaponry. Just this June, Bloomberg News obtained a document showing the Department of Defense had revised its guidelines transitioning away from human-in-the-loop systems, in which humans initiate actions assisted by AI, toward human-on-the-loop systems, in which AI can initiate actions under human supervision. China, which has been developing its own AI-enabled weaponry, has endorsed MHC but has stopped short of supporting the comprehensive ban on LAWS proposed by the UN Secretary-General. Instead, Beijing has advocated for a more limited international regulatory framework governing the development and use of autonomous weapons.

Three configurations of AI-related systems based on the degree of human involvement.

Ethical Frameworks

These disagreements have produced competing proposals for how autonomous weapons should be governed. States, international institutions, and industry professionals have all weighed in on the debate, recommending a variety of different ways to regulate the emerging technologies. Some countries have argued that the already existent International Humanitarian Law (IHL) is sufficient for the governance of LAWS. IHL includes a set of rules designed to govern armed conflict including measures intended to prevent indiscriminate attacks. The principles, which come from a series of treaties and customary laws, already bind nations to conduct that in theory should prohibit ‘unethical’ uses of LAWS. However, others, such as the International Committee of the Red Cross (ICRC), emphasize that IHL was designed specifically with human operators in mind and contend that further concrete frameworks are needed to regulate LAWS.

The position of UN member states on a convention banning LAWS, according to an analysis by Statista. China, which supports a ban according to this map, maintains a nuanced position.

Beyond legal frameworks, the geospatial community has also developed ethical guidance relevant to the use of GeoAI. For example, the LOCUS Charter, developed by EthicalGEO and Benchmark Initiative, sets out principles for the responsible use of location data. Although much of the charter focuses on issues such as privacy, surveillance, and data governance, several principles are applicable to GeoAI in autonomous weapons. In particular, Principle 2, Understand Impacts, emphasizes that users of location data should fully understand the potential consequences of its use before making decisions. Principle 3, Do No Harm, further states that location data should not be used in ways that discriminate, exploit, or cause unnecessary harm. While the charter wasn’t written with military applications in mind, its principles could offer policymakers a useful lens for thinking about how GeoAI in warfare should be regulated and used.

Another possible source of ethical guidance comes from AI developers themselves. Anthropic, the company behind Claude, was until recently involved in Project Maven, one of many partnerships between commercial AI companies and the U.S. military on GeoAI applications. However, the company has also publicly identified ethical limits on military uses of its models. In a statement on its defense partnerships, Anthropic stated that its AI systems “cannot be relied upon to exercise the critical judgment that our highly trained, professional troops exhibit every day,” and therefore should not be used to power fully autonomous weapons. While Anthropic supports the use of AI, including GeoAI, for military applications such as intelligence analysis and decision support, it has drawn an ethical red line at systems that make lethal decisions without meaningful human judgment.

Toward International Regulation

The United Nations has tried to bring all of these perspectives together as the organization works to chart a clear path forward on GeoAI in warfare. However, translating broad agreement on the risks of autonomous weapons into international regulation has proven far more difficult. Secretary-General António Guterres has opposed LAWS since 2018, and in the 2023 New Agenda for Peace, he argued that only a legally binding international instrument could effectively prohibit weapons systems that function without meaningful human control. His deadline for this treaty, which some have compared to international bans on other weapons, was set at 2026. That year has arrived, but the treaty has not.

References

Akella, Mamata. “What’s GeoAI? How AI Enhances GIS and Geospatial Data Analytics.” Felt, December 23, 2025. https://felt.com/blog/what-is-geoai.

Agbeja, Ayo Ore. “Human In, On, and Out of the Loop: Designing the Right Role for People in AI Systems.” Medium, September 1, 2025. https://medium.com/@Ayo.ore/human-in-on-and-out-of-the-loop-designing-the-right-role-for-people-in-ai-systems-ce2f51e675fc.

Amodei, Dario. “Statement from Dario Amodei on Our Discussions with the Department of War.” Anthropic, February 26, 2026. https://www.anthropic.com/news/statement-department-of-war.

Benchmark Initiative and EthicalGEO. The Locus Charter. March 2021. https://ethicalgeo.org/wp-content/uploads/2021/03/Locus_Charter_March21.pdf.

BSMA Enterprises. “Edge AI + UAV for On-Site Decision Making: Processing Data Direct.” BSMA Insight, June 24, 2025. https://bsmaenterprises.com/insights/edge-ai-uav-for-on-site-decision-making-processing-data-direct/.

Dayal, Umang. “Geospatial Data for Autonomous Defense Systems.” Digital Divide Data. Accessed July 27, 2026. https://www.digitaldividedata.com/blog/geospatial-data-for-autonomous-defense-systems.

Department of War. “DOW Unleashes ‘Agent Network’ to Transform AI-Enabled Battle Management and Targeting.” News Release, June 25, 2026. https://www.war.gov/News/Releases/Release/Article/4526862/dow-unleashes-agent-network-to-transform-ai-enabled-battle-management-and-targe/.

Erwin, Sandra. “Geospatial Intelligence Agency Urges Faster AI Deployment.” SpaceNews, May 20, 2025. https://spacenews.com/geospatial-intelligence-agency-urges-faster-ai-deployment/.

Fleck, Anna. “Twelve Countries Say No to Banning Autonomous Weapons.” Statista, January 27, 2026. https://www.statista.com/chart/35745/state-positions-on-legally-binding-autonomous-weapons-systems-ban/. 

Guterres, António. “Secretary-General’s Remarks to the Opening of the First Global Dialogue on Artificial Intelligence Governance.” United Nations, Geneva, Switzerland, July 6, 2026. https://www.un.org/sg/en/content/sg/statements/2026-07-06/secretary-generals-remarks-the-opening-of-the-first-global-dialogue-artificial-intelligence-governance-delivered.

Helene Cooper, “After U.S. Strike on Iranian School, Months Pass Without Answers,” The New York Times, June 16 2026, https://www.nytimes.com/2026/06/16/us/politics/us-strike-iranian-school.html.

Hunter, Lance Y., and Craig Albert. “The Military Application of Artificial Intelligence Technology in the United States, China, and Russia and the Implications for Global Security.” Defense and Security Analysis 39, no. 2 (2023): 207–232. https://doi.org/10.1080/14751798.2023.2210367.

International Committee of the Red Cross. “Joint Call by the United Nations Secretary-General and the President of the International Committee of the Red Cross for States to Establish New Prohibitions and Restrictions on Autonomous Weapon Systems.” News Release, October 5, 2023. https://www.icrc.org/en/document/joint-call-un-and-icrc-establish-prohibitions-and-restrictions-autonomous-weapons-systems.

International Committee of the Red Cross. “Three Lessons on the Regulation of Autonomous Weapons Systems to Ensure Accountability for Violations of IHL.” ICRC Law & Policy Blog, March 2, 2023. https://blogs.icrc.org/law-and-policy/2023/03/02/three-lessons-autonomous-weapons-systems-ihl/.

International Committee of the Red Cross. “UN Security Council: We Cannot Let AI Be Deployed on the Battlefield without Oversight and Regulation.” Statement, September 26, 2025. https://www.icrc.org/en/statement/we-cannot-let-AI-be-deployed-on-battlefield-without-oversight-and-regulation.

Kent, Jo Ling, Eleanor Watson, and Joe Walsh. “Pentagon Formally Designates Anthropic a Supply Chain Risk amid Feud over AI Guardrails.” CBS News, updated March 5, 2026. https://www.cbsnews.com/news/pentagon-anthropic-supply-chain-risk-feud-ai-guardrails/.

Mako, Gerald. “Human Oversight with Chinese Characteristics: Lethal Autonomous Weapons in the CCW GGE.” Articles of War, March 27, 2026. https://lieber.westpoint.edu/human-oversight-chinese-characteristics-lethal-autonomous-weapons-ccw-gge/.

Manson, Katrina. “Pentagon Sees Bigger Role for AI in Setting Military Targets.” Bloomberg, June 25, 2026. https://www.bloomberg.com/news/articles/2026-06-25/pentagon-sees-broader-role-for-ai-in-setting-military-targets.

National Geospatial-Intelligence Agency. “GEOINT Artificial Intelligence Strategy.” National Geospatial-Intelligence Agency. Accessed July 27, 2026. https://www.nga.mil/news/GEOINT_Artificial_Intelligence_.html.

National Geospatial-Intelligence Agency. “NGA Prepares to Use Generative AI for Mission Support.” National Geospatial-Intelligence Agency. Accessed July 27, 2026. https://www.nga.mil/news/NGA_prepares_to_use_Generative_AI_for_mission_supp.html.

Sayler, Kelley M. Defense Primer: U.S. Policy on Lethal Autonomous Weapon Systems. CRS In Focus IF11150. Washington, DC: Congressional Research Service, March 26, 2026. https://www.congress.gov/crs-product/IF11150.

Simmons, Adam. “The New Battlespace: How Geospatial AI Is Reshaping Military Intelligence.” Project Geospatial, March 12, 2026. https://projectgeospatial.org/geospatial-frontiers/the-new-battlespace-how-geospatial-ai-is-reshaping-military-intelligence.

People’s Republic of China. Working Paper of the People’s Republic of China on Lethal Autonomous Weapons Systems. CCW/GGE.1/2022/WP.6. Submitted to the Group of Governmental Experts of the High Contracting Parties to the Convention on Certain Conventional Weapons, July 2022. https://documents.unoda.org/wp-content/uploads/2022/07/Working-Paper-of-the-Peoples-Republic-of-China-on-Lethal-Autonomous-Weapons-Systems%EF%BC%88English%EF%BC%89.pdf. 

Trumbull, Charlie. “Considering a Legally Binding Instrument on Autonomous Weapons.” Lawfare, October 28, 2024. https://www.lawfaremedia.org/article/considering-a-legally-binding-instrument-on-autonomous-weapons.

United Kingdom Government. “The Bletchley Declaration by Countries Attending the AI Safety Summit, 1–2 November 2023.” Policy Paper. Updated February 13, 2025. https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration/the-bletchley-declaration-by-countries-attending-the-ai-safety-summit-1-2-november-2023.

United Nations Office for Disarmament Affairs. “Lethal Autonomous Weapon Systems.” United Nations Office for Disarmament Affairs. Accessed July 27, 2026. https://www.unoda.org/en/our-work/emerging-challenges/lethal-autonomous-weapon-systems.

United States of America. Humanitarian Benefits of Emerging Technologies in the Area of Lethal Autonomous Weapon Systems. CCW/GGE.1/2018/WP.4. Submitted to the Group of Governmental Experts of the High Contracting Parties to the Convention on Certain Conventional Weapons, April 3, 2018. https://ogc.osd.mil/Portals/99/Law%20of%20War/Practice%20Documents/US%20Working%20Paper%20-%20Humanitarian%20benefits%20of%20emerging%20technologies%20in%20the%20area%20of%20LAWS%20-%20CCW_GGE.1_2018_WP.4_E.pdf.

Varenikova, Maria. “How Ukraine Uses A.I. to Knock Deadly Russian Drones Out of the Skies.” The New York Times, June 15, 2026.  https://www.nytimes.com/2026/06/15/world/eu