An ideal zero ohm resistor would have zero power dissipated in it for any finite current through it, on the other hand it would have infinite power dissipated for any finite voltage across it.
Ideal zero ohm resistors don't exist in the real world though. A real "zero-ohm" resistor will have some resistance.
Resistor datasheets usually cover a whole series of resistors. Distributors use these datasheets to fill in the parametrics on their websites. However this can lead to parametrics that are misleading and/or nonsensical, particularly for more extreme components in the range. To Digi-Keys credit in this case they have filled in the tolerance column with "jumper" rather than a meaningless percentage.
Generally a series of resistors will have a maximum power rating, but it may also have maximum voltage and/or current ratings. For medium resistances the power rating is normally the limiting factor, but for high-value resistors the voltage rating is often the limiting factor and for very low-value resistors the current rating may be the limiting factor.
The datasheet specifies a maximum resistance of 20mΩ for the jumper. If power rating was the only constraint one could use this to calculate a maximum safe current.
$$ P = I^2R$$
$$ 0.1 = I^2 0.02 $$
$$ I^2 = 5 $$
$$ I \approx 2.24 $$
However the row in the table also says "Jumper, Imax. = 2.0 A". To me that says you shouldn't go over 2A, even though the power rating would theoretically allow for slightly more.