There are various options for RNA polymerases commonly used in vitro [1], generally the error rate is considered to be "pretty low", something like 1 in 10k. An

There are various options for RNA polymerases commonly used in vitro [1], generally the error rate is considered to be "pretty low", something like 1 in 10k. An error will also have to result in a change of amino acid, which less likely due to the redundancy of the code. So any error mRNA will probably only have one amino acid changed. This could cause the resulting spike protein to aquire some new characteristic, but it's not very likely I would guess. Also, in the clinical trials they test the actual produced vaccine, with all errors included. [1] https://international.neb.com/tools-and-resources/selection-charts/rna-polymerase-selection-chart https://international.neb.com/tools-and-resources/selection-...