IHL, in particular its rules governing the conduct of hostilities (i.e. the manner in which a war is fought), are relevant insofar as the international community has not prohibited certain means or methods of warfare. Currently, no automated or autonomous technology is banned, although states have been under political, scientific, and civil society pressure to ban fully autonomous lethal weapons since the launch of the Ban Killer Robots movement2. In 2018, for example, the European Parliament adopted a resolution calling on the European Commission, member states and the European Council to „work towards the opening of international negotiations on a legally binding instrument banning lethal autonomous weapons systems“ (European Parliament 2018, paragraph 3). In the absence of such a treaty, existing rules of international humanitarian law govern its application. The most likely starting point for international legal regulation in this direction would be the European Union, as it is the only international organisation with institutional capacity, a pre-existing mandate and the political will to adopt binding rules of such scope (in fact, it has already taken the first steps towards intra-Community regulation of AI; see European Commission 2020). Although the European Union officially only has the power to legislate against its member States, the effect of any resulting regulation would extend beyond the organization`s borders. The situation could be similar to the European Union`s General Data Protection Regulation – to the extent that foreign companies offering automation products and services want to operate in the EU market, they would be obliged to comply with applicable EU rules. This is a strategic opportunity for the European Union, as it is particularly well placed to serve as a pioneer in this field, thus guiding the debate on the appropriate legal and regulatory regime for autonomous technologies.

New technologies pose normative challenges for both national and international law, in particular with regard to the adequacy of existing regulations. Some technology-specific issues will inevitably arise and require regulatory and legislative action. The resulting normative development will first take place in the domestic environment, as international legislation is a relatively slow process, especially in areas where national security is linked. Press Association. 2019. „Mark Zuckerberg calls for stricter regulation of the internet.“ The Guardian, March 30. www.theguardian.com/technology/2019/mar/30/mark-zuckerberg-calls-for-stronger-regulation-of-internet. International law, on the other hand, will largely play the role of a bystander when it comes to stand-alone commercial solutions for civilian purposes. However, the international community may at some point feel the need to harmonize countries` national laws to ensure that domestic legal regulation of these commercial technologies is consistent across borders. The legal mechanism for harmonization would be the adoption of a so-called law harmonization treaty, which obliges States parties to the instrument to adopt national laws relating to their criminal, civil or administrative law.

For example, such an agreement could impose uniform security standards, liability rules, certification systems, data management processes, requirements for human oversight of the use of technology, failover mechanisms to be established, operational constraints, rules on bias and criminal offences related to autonomous technologies. A more practical challenge is that it`s hard to regulate something you don`t fully understand. Autonomous technologies are still in their infancy and it is difficult, if not impossible, to predict scientific developments in this field, even in the short term. All new laws and regulations should be broad enough not to quickly become obsolete, but also not so vague that they do not provide meaningful guidance. The difficulty of this undertaking inevitably means that no comprehensive set of sectoral rules will be adopted in the near future – neither national legal acts for autonomous technologies, nor an international treaty on autonomy as such. Instead, we can expect discrete rules that govern relatively specific aspects of autonomous technologies. Since war is governed by a dense international legal framework, there are already many rules governing the use of autonomous technologies in wartime. These rules form a regime of international law known as international humanitarian law (IHL), also known as the law of armed conflict. The advent of new technologies always raises questions about their legality, and this certainly applies to autonomous technologies, including those that use varying degrees of artificial intelligence (AI). When developing and implementing stand-alone solutions, countries must ensure that their use is consistent with established moral and ethical principles, which are often enshrined in national and international legislation. The fundamental legal dilemma of any new technology is whether existing law is capable of regulating it in accordance with these principles and, if not, what new legal instruments are needed to achieve this objective. This article explores this question in the context of autonomous technologies.