JupiteX Get the app
Defence & Security16 Nov 2026 · about 7 min

UN chief urges countries to rein in ‘merciless’ killer robots

The brief

Lethal autonomous weapons systems are weapons that can select and attack targets without a person deciding on each individual strike. The article calls them “killer robots” and says armies are exploring AI to identify and strike targets more quickly and accurately. The central concern is who—or what—makes life-and-death choices during combat. In general, such systems use sensors and software to process information about their surroundings, identify possible targets, and act according to their design. Human involvement can vary: a person may approve a strike, or the system may act after activation. The article does not explain the specific technical steps or control arrangements of any weapon. That distinction matters because autonomous systems could make lethal choices at speed, while people may have limited time to intervene. UN chief Antonio Guterres is urging countries to negotiate binding rules, including clear prohibitions and restrictions. The ICRC is calling for bans on unpredictable systems and those designed or used to target humans.

01

What are lethal autonomous weapons systems, and how do they choose and attack targets?

Lethal autonomous weapons systems are weapons that can select and attack targets without a person deciding on each individual strike. The article calls them “killer robots” and says armies are exploring AI to identify and strike targets more quickly and accurately. The central concern is who—or what—makes life-and-death choices during combat.

In general, such systems use sensors and software to process information about their surroundings, identify possible targets, and act according to their design. Human involvement can vary: a person may approve a strike, or the system may act after activation. The article does not explain the specific technical steps or control arrangements of any weapon.

That distinction matters because autonomous systems could make lethal choices at speed, while people may have limited time to intervene. UN chief Antonio Guterres is urging countries to negotiate binding rules, including clear prohibitions and restrictions. The ICRC is calling for bans on unpredictable systems and those designed or used to target humans.

02

What binding rules is the UN chief asking countries to negotiate?

UN chief Antonio Guterres is asking countries to negotiate a legally binding instrument to regulate autonomous weapon systems. He says the rules should include clear prohibitions and restrictions. The aim is to prevent the use and spread of weapons he described as “merciless,” and to protect civilians from the devastation of armed conflict.

The proposed instrument would set binding rules rather than rely only on voluntary promises. The article does not provide a draft treaty or list every rule it might contain. The ICRC has called for bans on two categories: “unpredictable autonomous weapons” and weapons designed or used to directly target humans.

At a conference in Geneva, representatives from 128 countries are deciding whether to open negotiations on binding rules. Guterres urged them to seize the moment, but agreement to begin talks this week appeared unlikely to observers. The United States has proposed a five-year process for a non-binding instrument, while NGOs and the ICRC advocate a one-year timeline.

03

How many countries are taking part in the UN meeting, and how did countries vote on starting negotiations?

The Geneva meeting brings together representatives of 128 countries. Their task is to decide whether to open negotiations on legally binding rules for autonomous weapon systems. This is a separate count from the vote at the UN General Assembly the previous month, where states considered a resolution calling for negotiations.

In that General Assembly vote, 163 states adopted the resolution. Five voted against: the United States, Russia, Belarus, Israel, and North Korea. Nine abstained, including Iran, Saudi Arabia, and China. The resolution called for negotiations on rules governing autonomous weapons; it was not itself a binding treaty.

The article says many states support tighter rules, while countries with the technology are reluctant to limit their options. Washington has proposed developing a non-binding instrument over five years; NGOs and the ICRC want a one-year timeline. Observers cautioned that agreement to start binding-rule talks at the Geneva meeting looked unlikely.

04

Why do militaries want to use AI-powered weapons on the battlefield?

The article gives a direct reason: armies are seeking to use artificial intelligence to boost their fighting abilities. They hope it will help identify and strike targets more quickly and accurately. In a battlefield where decisions may need to happen fast, those capabilities can appear militarily useful.

The article points to competing efforts involving drones. Russia has reportedly used fully autonomous drones with deadly effect in Ukraine, while Ukraine’s defence industry is pushing to develop autonomous “drone swarms.” These examples show why AI-enabled weapons are part of current military planning, although the article does not describe their technical performance or how individual strikes are authorized.

The pursuit of these capabilities has intensified calls for limits. Guterres urged countries to negotiate binding rules, and the ICRC said AI developments have made regulation more urgent. But states that possess the technology may resist rules that constrain its use. The meeting’s immediate decision is whether to begin negotiations, not whether a treaty has already been agreed.

05

What could happen to civilians if weapons select and strike targets without meaningful human control?

If weapons select and strike targets without meaningful human control, civilians could be exposed to lethal decisions that are difficult to challenge or stop. The article does not describe a specific civilian attack caused by such a system, but it reports warnings about the risks of handing life-and-death power to autonomous machines.

Human Rights Watch said civilians would be “dangerously exposed” if the UN meeting failed to act decisively. Guterres urged countries to prevent the use and spread of these weapons, saying civilians must be shielded from the carnage and devastation of war and from unnecessary suffering. The ICRC also seeks bans on unpredictable systems and those designed or used to target humans.

The danger is a concern raised by campaigners and humanitarian groups, not a confirmed outcome for every autonomous weapon. The meeting’s 128 countries are considering whether to open negotiations on binding rules. If rules are negotiated, they could establish prohibitions or restrictions; observers cautioned that even agreeing to start talks this week looked unlikely.

06

What rules already govern how weapons may be used in war, and how could a new treaty change them?

The article names the Convention on Conventional Weapons, whose Review Conference is hosting the debate, but it does not detail the existing rules that govern attacks. More broadly, established international humanitarian law requires parties to distinguish civilians from combatants, avoid attacks expected to cause excessive civilian harm, and take feasible precautions. These are general legal principles, not details supplied by the article.

The proposed new instrument would address autonomous weapon systems specifically. Guterres called for a legally binding instrument with clear prohibitions and restrictions. The ICRC has proposed banning unpredictable autonomous weapons and systems designed or used to directly target humans. These measures would add specific limits rather than simply repeat the article’s discussion of the existing convention.

Countries at the Geneva meeting are deciding whether to open negotiations, not adopting a treaty there. At the General Assembly, 163 states backed a resolution calling for negotiations, while five opposed it and nine abstained. If negotiations begin and states agree on a treaty, it could set binding limits; the article says agreement to start talks this week looked unlikely.

07

How does artificial intelligence let a machine identify objects and make decisions from data?

The article does not explain how artificial intelligence works. In general, many AI systems are trained on examples: software adjusts its internal settings to detect patterns in data, such as images or sensor readings. When it receives new information, it uses those learned patterns to classify what it is seeing or estimate what action may fit a programmed goal.

For a weapon, that could mean processing sensor data to identify an object as a possible target and then following rules about whether to act. This is a general explanation, not a description of any particular weapon in the article. The source says only that armies want AI to help identify and strike targets more quickly and accurately.

AI does not make every decision in the same way, and the degree of human control can vary by system. That matters because the article’s debate is about whether autonomous weapons should be restricted or prohibited. Guterres is urging countries to negotiate binding rules, while the ICRC calls for bans on unpredictable systems and those designed or used to target humans.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web