Last updated, 04/10/2025

Artificial Intelligence in conflict and public international law

According to the UN, “the human rights framework provides an essential foundation that can provide guardrails for efforts to exploit the enormous potential of AI, while preventing and mitigating its enormous risks (source: UN High Commissioner for Human Rights).

Considering AI is such a new field of technology, law may at times struggle to keep up as the technology further develops and improves. It is still not clear whether international human rights law should be applicable when undertaking assessments of the legality of AI capabilities aimed to be used in conflict. Certain countries, such as the United kingdom, are currently leading the discussions on this subject (source: ICRC Law and Policy Blog). 

It could be argued that because the human rights framework is applicable at all times, this essentially means that it should also apply, at least to some extent, to the legality assessments of AI capabilities. This is further supported by the commentary under the customary law’s Fundamental Guarantees and Rule 87 relating to the principle of humanity.

However, it should be emphasised that during conflict the international humanitarian law (IHL) is the default legal basis/framework that takes precedence over international human rights law. In such situations, the role of the human rights framework is to supplement IHL where any gaps exist in such circumstances. In addition, the applicability of human rights protections depends on whether the individual concerned is within the “jurisdiction” of the State involved.

Lastly, there is one major difference between the applicability of human rights and IHL. Under human rights law all persons are equally protected against any breaches. However, under IHL combatants and civilians are not afforded equal protection against attacks. Force can be used against combatants unless they lay down their arms, although such force should not be superfluous (source: ICRC International Humanitarian Law, a Comprehensive Introduction, p. 35, 07/2022).

In terms of human rights instruments that are truly AI focused, there have been two recent developments in Europe in this respect.

The first international treaty on AI – the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law* – was introduced in 2024. The treaty unfortunately lacks enforcement mechanisms, and moreover, it does not apply to defence and national security matters. Secondly, the updated data protection Convention 108+ has recently been supplemented with the addition of the AI Guidelines. These guidelines too are non-binding as per their legal status as advisory instruments rather than legal instruments.

Finally, there is the EU AI Act (the “AI Act”) which is yet to come fully into force. The AI Act has an extraterritorial scope, applying to organisations that place AI systems on the EU market, put them into service in the EU, or whose AI output is used in the EU, regardless of the organisation’s location. This means non-EU companies must comply if their AI targets EU users or produces output intended for use within the EU. The AI Act makes exceptions for certain uses, such as AI in national security, defence, and some research activities. However, the defence/military use does become relevant in the context of the AI Act if a system has a dual military and civilian purpose, as well as in the military/defence procurement stages, including during the development stages by private companies with the use intended for military/defence purposes. Although not a human rights instrument per se, the AI Act provides for a mechanism to assess high risk AI systems from the human rights law angle. Since AI systems can lead to biases and discrimination, the AI Act aims to safeguard against such negative outcomes by imposing strict requirements on high-risk AI systems that can affect fundamental rights of individuals.

Relevant Laws

The current (pre-AI) human rights framework consists of: 

(i) regional treaties agreed at a regional level, of which the most notable are:

• the European Convention on Human Rights (ECHR),
• the African Charter on Human and Peoples’ Rights,
• the American Convention on Human Rights, or
• the ASEAN Human Rights Declaration

(ii) state laws about human rights

• e.g. in the UK, this is principally the Human Rights Act 1998, which incorporates the ECHR

( iii) global treaties by international bodies

• 9 core treaties principally established by the United Nations, listed here.
  

The ratification status of a treaty is nevertheless important as it indicates whether a treaty has been incorporated into local state law. In other words, whether it is binding on the state in question.

* The Framework Convention does not apply to national defence matters nor to research and development activities, except when the testing of AI systems may have the potential to interfere with human rights, democracy, or the rule of law.