During the event Def Con held last August some attack simulations were carried out against the terminals that will be used in the USA 2020 Presidential Elections. The results of the tests, conducted on a hundred stations, would have highlighted different problems from the point of view of safety with each machine exposed to at least one potential violation methodology.
The simulations were carried out by researchers at Georgetowen University (Washington DC) who would have already informed the US institutions of the risks to the democratic process, the new appointment with the vote at the federal level is expected for November next year and before then it will be necessary to make the necessary corrections (if possible).
As anticipated, it would not have been a single vulnerability common to all the devices, possibly solved with a single specific intervention, but more critical ones, including some particularly simple problems such as the use of passwords too "weak" or failure to apply of algorithms for encryption.
As regards the possibilities of illegal access to the system, it could be carried out both remotely and locally, this and some recent episodes that have questioned the reliability of electronic voting could even lead to an exclusion of this method in favor of more traditional approaches such as paper voting.
In this regard it would be enough to cite the case of the State of the Georgia who would decide to repeal electronic voting following a not sufficiently transparent management by the governor (and candidate) Brian Kemp. I note his refusal to accept the appropriations offered by the federal government for the upgrade of the voting terminals.