The European Parliament adopted by 364 votes to 274, with 52 abstentions, a resolution on artificial intelligence (AI): questions of interpretation and application of international law in so far as the Union is affected in the areas of civil and military uses and of state authority outside the scope of criminal justice.
EU framework on AI and guiding principles
Parliament called for the adoption of a common European legal framework with harmonised definitions and common ethical principles, including for the use of AI for military purposes. AI for defence purposes should be accountable, fair, traceable, reliable and governable. In all cases, technologies should be developed in a safe and technically rigorous manner.
Moreover, AI is a scientific advance which should not undermine the law but should, on the contrary, always be governed by it. Under no circumstances should AI, robotics and related technologies violate fundamental rights, democracy and the rule of law.
Members recalled that AI cannot replace the human being in the judicial process when it comes to making judgements or any final decision. Thus, when using evidence emanating from technologies using AI, judicial authorities should be obliged to give reasons for their decisions.
International law and military uses of artificial intelligence
Parliament stressed that the use of AI is an opportunity to enhance the security of the European Union and its citizens and that it is essential for the Union to adopt a coherent approach in the forthcoming debates on this subject at international level. It invited the EU to take the lead and play an active role, together with the United Nations and the international community, in promoting a global framework governing the use of AI for military and other purposes, ensuring that such use respects the strict limits set by international law and international humanitarian law.
Parliament stressed that AI used in a military and a civil context must be subject to meaningful human control, so that at all times a human has the means to correct, halt or disable it in the event of unforeseen behaviour, accidental intervention, cyber-attacks or interference by third parties with AI-based technology or where third parties acquire such technology.
Autonomous decision-making should not absolve humans from responsibility, and that people must always have ultimate responsibility for decision-making processes so that the human responsible for the decision can be identified.
Lethal autonomous weapon systems (LAWS)
Members recalled that Parliament called for the drafting and urgent adoption of a common position on lethal autonomous weapon systems, preventing the development, production and the use of LAWS capable of attack without meaningful human control, as well as the initiation of effective negotiations for their prohibition. It insisted on the need for an EU-wide strategy against LAWS and a ban on so-called killer robots.
The use of lethal autonomous weapon systems raises fundamental ethical and legal questions about the ability of humans to control these systems. Such systems should meet a minimum set of requirements and be used as a last resort. They should only be considered lawful if they are subject to strict human control.
State authority: examples from civil areas, including health and justice
Parliament urged the Member States to assess the risks related to AI-driven technologies before automating activities connected with the exercise of state authority, such as the administration of justice. It invited the Commission to assess the consequences of a moratorium on the use of facial recognition systems until the technical standards can be considered fully fundamental rights-compliant and that there are strict safeguards against misuse.
Members expressed serious concerns about some highly intrusive social scoring applications that have been developed, as they seriously jeopardise respect for fundamental rights. They called for an explicit ban on the use of mass social scoring (for monitoring and rating citizens) by public authorities as a means of restricting citizens' rights.
Moreover, given that AI is called upon to play an increasingly fundamental role in health matters, in particular through diagnostic aid algorithms, robot-assisted surgery and intelligent prostheses, Parliament insisted that all uses of AI in public health matters should respect the protection of patients' personal data and avoid the uncontrolled dissemination of such data.
Transport
Members took note of the significant economic potential of AI applications in this area. They stressed the need to promote AI to foster the multimodality, interoperability and energy efficiency of all modes of transport, including in the field of military logistics.
They also stressed that the circulation of autonomous vehicles in the European Union, which is likely to lead to a particularly high number of disputes under international private law, should be the subject of specific European rules stipulating the legal regime applicable in the event of cross-border damage.