This paper presents an electromagnetic-artificial intelligence design optimization approach of a wireless power transfer link (WPTL). Electromagnetic solutions from a 3-DFE model of the WPTL were utilized to minimize the field signature around each design while maintaining high efficiency. The reduced field signatures around the WPTL would enable the achievement of designs that are in accordance with international electromagnetic compatibility (EMC) standards such as international commission on nonionizing radiation protection guidelines with respect to public exposure levels of EM signatures below 27 μT. As the optimization process involved utilizing a large number of fitness function evaluations, numerous 3-DFE solutions were required increasing the computational burden. Artificial neural networks were developed, trained, and used to produce the required equivalent 3-DFE solutions to evaluate the fitness function. As a result, this process significantly reduced the computational time by nearly 90%. The genetic algorithm-based optimization process yielded the desired results of electromagnetically compatible WPTL designs at the early development stage satisfying EMC standards.