Free beta access. 50 free AI tokens each week. Optional top-ups available. See what is included
Opening Word document…

European nations agree on the minimal principle of Human Control.

Nadibaidze, 2021 - Ph.D. Student at the University of Southern Denmark [Anna Entry submitted for the Second OSCE-IFSH Essay Competition “Commitment to Control over Weaponised Artificial Intelligence: A Step Forward for the OSCE and European Security” https://www.osce.org/files/f/documents/8/3/507341.pdf Acc. 4/21/22 TA]

Second, the discourses surrounding weaponised AI – the ways that OSCE participating States talk about LAWS – also have considerable impact on European security. Both a common definition of LAWS and an agreement on the appropriate level of human control over weapons systems are lacking, which gives way for misinterpretation and increases security risks. Let us examine the discourses of three major players in European security: France, the Russian Federation, and the United Kingdom. Their official positions converge on the importance of retaining human control. The Minister of the French Armed Forces, Florence Parly, said that “France refuses to entrust the decision of life or death to a machine that would act in a completely autonomous manner and would be beyond any human control” (Ministère des Armées 2019). Russia said it “is committed to the need to maintain human control over LAWS, no matter how ‘advanced’ these systems may be” (Russian Federation 2021, 3). The UK Ministry of Defence noted in January 2021, “the operation of our weapon systems will always be under human control and no UK weapons systems will be capable of attacking targets without this” (Ministry of Defence 2021). Nevertheless, autonomy and the concept of appropriate human control over weapons systems are perceived differently by these States. The Russian side remains opposed to a legally binding treaty that would ban LAWS, arguing that the definition of LAWS should “strike a balance between humanitarian concerns and legitimate defence interests of States” (Russian Federation 2021, 3). France has suggested a division between “fully” and “partially” lethal autonomous weapon systems and adopting different types of measures for these two categories. In the French perspective, only the “fully” autonomous weapons should be prohibited (French Republic 2021). Meanwhile, the UK has stated that “an autonomous system is capable of understanding higher-level intent and direction”, a definition that is more precise and constraining than those of other States (Ministry of Defence 2017, 13). A UK House of Lords Select Committee report said this definition is “clearly out of step with the definitions used by most other governments” and also “hamstrings attempts to arrive at an internationally agreed definition”(Select Committee on Artificial Intelligence 2018, 105). As a common denominator, these States agree on the principle that weapons systems should not function completely autonomously. However, the differences in their perceptions hinder the progress on understanding and preventing the security risks related to the use of weaponised AI. They create misperceptions about the uses of AI, specifically between the leaders in this sphere, who are all carefully watching each other’s technological developments. The NSCAI, for instance, warned the US government that “competitors are actively developing AI concepts and technologies for military use,” specifically focusing on China and Russia (2021, 22). The discrepancies in definitions and discourses create risks of misunderstanding when, for instance, one participating State is developing a certain weapons system considered to be LAWS by another State. Such communication issues can lead to a security dilemma in which “one state’s pursuit of greater automation and faster reaction times undermines other states’ security and leads them to similarly pursue more automation just to keep up,” and encourgages experts to speak of an ‘AI arms race’ (Scharre 2021).

Sign in to recut, update, or save this card.