It's easy to see both sides of the decision to presume or not presume that machines work - we should like somehow to explain that on the one hand it's unlikely that without his intent Jim's iPhone sent Gemma a series of messages threatening to rape her, and so we shouldn't need to drag Apple engineers into court to convict Jim - but then on other hand it's far from certain that this bank database which says I made 590 cash withdrawals in sixteen minutes is correct, more likely a bug duplicated a single withdrawal I made, and we should require the bank to produce evidence otherwise if they insist I made these withdrawals.
I can imagine that smart lawyers motivated by some real world examples and assisted by independent software engineers could come up with a general formula for courts, something a bit larger than either the old rule or the new one, but still relatively pithy. Whether a government would ever make time to do that work is another question.
https://www.prospectmagazine.co.uk/ideas/law/64464/post-offi...
I found it fascinating to learn that UK law explicitly does (but did not always) encode this presumption:
https://emptycity.substack.com/p/computer-says-guilty-an-int...
(This link is to the substack of David Allen Green, an excellent legal commentator in the UK.)