I'm certainly no expert in randomness generators but the notion has been floating around for awhile that quantum indeterminancy is the best option. Here's a paper on it:
> "The presented above examples clearly show that classic random number generators may be exposed to various attacks, or may have the so-called backdoors. This justifies the need to develop alternative technologies that could replace the classic generators on a large scale. The most promising, because they have a fundamental justification for the randomness in the formalism of quantum mechanics, are quantum random number generators."
https://www.nature.com/articles/s41598-021-95388-7
> "The presented above examples clearly show that classic random number generators may be exposed to various attacks, or may have the so-called backdoors. This justifies the need to develop alternative technologies that could replace the classic generators on a large scale. The most promising, because they have a fundamental justification for the randomness in the formalism of quantum mechanics, are quantum random number generators."