Artificial intelligence is now a major issue of security and military sovereignty for states. However, writes Rémi Meunier, AI really takes its value only in support of a human spirit, capable of exploiting information.
“How to deploy NATO forces in support of a government of a small country confined and weakened by a deadly epidemic and the emergence of armed groups.” This scenario, imagined by the North Atlantic Treaty Organization in 2018 as part of its “innovation” challenge takes , eighteen months tardlater, a particular resonance. At the time of the pandemic,, it is interesting to see what techniques have been selected by NATO to address this problem.
Unsurprisingly, artificial intelligence (AI) is one of the tracks chosen, especially in the context of “data filtering and fusion, visualization and predictive analysis”. analyse “. Thus, the analysis of cross-images with data of visits to restaurants or public transport allows to study the link between population movements and the spread of the disease.
Homeland Security and”Opex””
This data collection, which feeds an algorithm so that it then provides an analysis, is above all valuable in the field of intelligence.
In the field of internal security, in the face of the terrorist threat, AI is thus an excellent way to sort millions of Twitter messages, through semantic analysis, to identify signals of radicalization and trace back to Islamist or supremacist networks – depending on the threats to l’analyse states. While the use of this data raises an obvious question of individual freedom, it has nevertheless become a strategic resource.
In external security, AI-enriched tools have also become valuable logistical support for the military during Opex (external operations, such as Barkhane in Mali). While the deployment of AI still requires a complex infrastructure that can complicate the recovery of information, it is booming and should provide the military with better visibility of the forces involved. This is all the more valuable because modern fighting is more and more like urban guerrillas than open clashes, as in the last century.
Inter-state and public-private cooperation
While the use of these technologies is still the sovereignty of states, it is hoped that cooperation initiatives will develop, particularly at European level. The new President of the Commission and former German Defence Minister Ursula von der Leyen has called for a convergence of European military and technological means over the next five years.
Another increasing cooperation is that between the Department of Defense and AI projects led by private actors – companies. These are en indeed at the forefront of the development of these technologies, and the public authorities are en increasingly seeking, through contracts and partnerships,to benefit from them. There are indeed a large number of application areas common to the military and private sector: predictive maintenance, optimization of logistics flows, security of transmissions… and so many solutions that are developing in companies, and which can be put to use by the defense for its specific use.
Ai is n’en still a tool
Collecting and s’en l’information extracting information from data processed by AI algorithms is of value only if the user, company or active force, is able to follow through on the information provided by the machine. In the case of bank fraud, d’y for l’intelligence example, artificial intelligence is able to identify abuses but not to stop them. Only the en establishment of an appropriate team and process can address the problem. Because if AI can reveal and explain problems, it only takes on value as a tool in the service of human action.
Today AI is no longer a technological problem. The data, computational capabilities and algorithms are there. The whole issue is the ability of actors,public or private,civilian or military,to exploit the information issued by this AI to solve problems. If the collaboration between AI and is still in its infancy,the prospects are great and must be exploited,with respect for individual freedom and control of access to data; that only strong legislation can guarantee.
Rémi Meunier is Dataiku’s manager of and security practices for Europe.