There is no fundamental theoretical limit on the current. From the formula for the radiated power of a single wire derived in my answer here, the radiation resistance is given by:
$$R_{rad}=\frac {2P_{rad}}{I_0^2}=\frac { \eta \ }{4 \pi} \biggl[\frac 1{kl} \biggr(2kl(-Ci(2kl)+\ln(kl)+\gamma-1+\ln 2)+\sin(2kl)\biggr)\biggl]$$
From this you can see that this resistance increases monotonically with frequency $f\sim kl$ (check the plot of my linked answer above). So, in higher and higher frequencies, the voltage needed to supply a specific current increases. But, as long as you have the necessary supply, you can always create such currents. However, there is some point at which practically supplying such a voltage becomes impossible because of the technological limitations on such high power voltage supplies at such high frequencies. At some ridiculously high voltages, even the air surrounding the circuit experiences dielectric breakdown and everything falls apart (the actual limitations are far below such a ridiculous voltage). This is all assuming that you have no skin effect and no frequency dependence in the actual copper resistance of your wire, which introduces further limitations on the current. In an actual wire, the copper resistance increases with frequency, which means that the $E_d=\frac12RI_0^2$ heat created completely melts the wire.
Nonetheless, I don't have enough experience with such high voltage sources at those frequencies to give you any numbers on when the limitations actually become practically impossible to overcome.