UN Warns of Growing Risks as Artificial Intelligence Advances in Autonomous Weapons
The rapid integration of artificial intelligence into military operations is raising urgent questions about human control over warfare, prompting a United Nations research institute to call for stronger international cooperation and more effective regulatory safeguards.
Robin Geiss, director of the United Nations Institute for Disarmament Research (UNIDIR), warned on Monday at the UN headquarters in New York that technological advances are bringing the prospect of increasingly autonomous warfare closer. The central challenge, he stressed, is determining whether machines should be allowed to select human targets and make decisions that result in death.
His warning comes as artificial intelligence becomes increasingly embedded in military capabilities, from intelligence assessment and target identification to drone operations, cyberattacks and influence campaigns.
Faster military decisions raise escalation concerns
The growing use of AI in defence systems is transforming the speed at which military decisions can be made. However, this acceleration also creates risks, particularly when human operators have less time to assess information, challenge automated recommendations or intervene before an operation is carried out.
According to Geiss, errors and misinterpretations could have serious consequences, especially when automated systems operate in situations involving uncertainty or heightened geopolitical tensions.
The risks extend beyond conventional battlefields. In nuclear-related scenarios, decisions made under severe time pressure could contribute to miscalculation and escalation, making the relationship between automation and human judgment a major international security concern.
The debate therefore goes beyond technological innovation. It raises fundamental questions about accountability, the limits of machine autonomy and the responsibility of states when AI-supported military systems cause harm.
UNIDIR expands international cooperation initiatives
In response to these challenges, UNIDIR has established a Centre of Excellence on AI, Peace and Security in Geneva. The initiative is intended to strengthen expertise and support international discussions on the implications of artificial intelligence for conflict prevention, disarmament and global stability.
The institute is also preparing several initiatives in New York to encourage dialogue among governments, technology companies, researchers and other stakeholders.
Among the planned measures are a multistakeholder forum, a voluntary framework promoting responsible conduct by technology companies, and a common glossary designed to clarify terminology surrounding military autonomy and security.
Establishing shared definitions is an important part of the process, as governments and technical experts need a common vocabulary to discuss the capabilities, limitations and risks of increasingly sophisticated military systems.
Non-state actors present another security challenge
UNIDIR is also examining the potential misuse of artificial intelligence by non-state armed groups, criminal networks and terrorist organisations.
The spread of accessible AI technologies could create additional security challenges if such actors exploit them to support malicious activities. Addressing these threats requires international coordination alongside efforts to develop responsible practices for military and civilian technologies.
The institute also plans to contribute to preparations for the Geneva AI Summit scheduled for June 2027, providing another opportunity to address the relationship between emerging technologies, peace and international security.
Disarmament Week puts military AI under scrutiny
Geiss's remarks come ahead of the United Nations Disarmament Week, observed from October 24 to 30. Established in 1995, the annual initiative seeks to improve public awareness and understanding of disarmament issues and their broader significance.
The growing military use of artificial intelligence adds urgency to these discussions, particularly as technological development continues to outpace international efforts to establish common safeguards.
For the United Nations, the challenge is to encourage innovation without allowing the increasing autonomy of weapons systems to undermine human responsibility, international stability and the protection of civilians.
As governments and technology companies explore new AI capabilities, the debate over how much authority machines should exercise in armed conflict is becoming an increasingly important issue for global diplomacy.
-
15:00
-
14:40
-
14:20
-
14:00
-
13:45
-
13:30
-
13:15
-
13:00
-
12:45
-
12:30
-
12:15
-
12:00
-
11:45
-
11:30
-
11:15
-
11:00
-
10:45
-
10:30
-
10:27
-
10:23
-
10:22
-
10:19
-
10:15
-
10:00
-
09:45
-
09:30
-
09:15
-
09:00
-
08:50
-
08:39
-
08:20
-
23:30
-
23:00
-
22:30
-
22:10
-
22:00
-
21:30
-
21:00
-
20:30
-
20:00
-
19:40
-
19:20
-
19:00
-
18:40
-
18:20
-
18:00
-
17:40
-
17:20
-
17:00
-
16:40
-
16:20
-
16:00
-
15:40
-
15:20
