Questions, but also and above all technical fears : for example, one sms gateway iraq wonders if and . How some malicious person could deliberately "Plagiarize" an ai, that is, be able to maliciously . Modify its behavior so as to make it make wrong decisions on purpose; and if . So, how it is possible to prevent or at least detect such an attack on . An ai, which by definition is "Unpredictable" in its reasoning. The debate is open, and . It is all the more important as the scope of ai application shifts towards critical .
Areas in which incorrect behavior could also compromise the lives of human beings: healthcare, defense, . Management of critical infrastructures and so on.Speaking of critical infrastructures, the topic is certainly not . New in itself but will certainly remain crucial in the coming years due to the . Ever-increasing use of connected ict technologies even where historically they were not previously applied. Naturally, . The wave of iot devices, as cheap as they are insecure, that is invading our . Homes, our companies, our factories and our industrial plants, certainly does not help us feel .
Safe for the future ; not to mention the upcoming g , which brings with . It both extraordinary application potential and the risk of unprecedented abuses and violations. However, it . Is not, as many think, just a problem of "Sovereign technology": embargoes against that supplier . Or that nation are not effective in themselves, and above all they constitute a political . Response to a technical problem; what should be done, and what europe is actually starting . To do albeit with culpable delay, is to establish clear and transparent technical rules to .
