The equation in the articles mentioned above can still be used for current measurements. Wherever voltage is mentioned, replace this with the current.
As an example, see the absolute accuracy or standard uncertainty calculation for the NI-9208 Current Input module below.
In the manual,
NI 9208 Datasheet, the Accuracy data is displayed in the
Specifications section, in a table.

Let's assume a measured value of 20 mA for this example.
(
VoltageReading x GainError)
To calculate the gain error, we multiply the
Percentage of Reading, here ± 0.76 %, by the measured value, in SI units as below.

(
VoltageRange x OffsetError)
Then calculate the offset error by multiplying the range, here 22 mA (from the footnote), by the
Percentage of Range, ± 0.04 %, as below.

Adding these two together will give ± 0.0001608 A or 160.8 µA. This is the absolute accuracy of the current module.
For your complete system accuracy, you will need to include all component absolute accuracy values, as shown in the equation below.
Additional Information
If the
System Noise needs to be included, the Root Mean Square (RMS) value needs to be converted to a peak value by dividing the RMS value by 0.7071 - keeping in mind the units - then you can add this to the absolute accuracy equation for your module.
Outside of the temperature ranges specified for your module (in the example here -40 °C to +70 °C), you will need to include the temperature coefficients under
Stability.