Embedding information in random bit sequences while maintaining certified randomness

[bsr]

https://arxiv.org/pdf/1911.00001

This is basically what we (OneRNG users) want to prevent and something that has concerned me for a while.  I'm posting this because it demonstrates working schemes for embedding data in a long random bit sequence at an information/random bit ratio > 1:100 and can still pass NIST randomness tests.  I keep wondering how many nefarious ways this could be used via adversarial sequences or data sets.

Reminds me of the number pi embedded in the attached image, but still passes dieharder.  Obvious to my eye but I applied an edge detect filter to make the repeating patterns "pop".  I only circled a couple obvious ones but lots of repetition in there.  The image is taken from this video: http://linux.conf.au/schedule/presentation/65/

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 649955 bytes
Desc: not available
URL: <http://lists.ourshack.com/pipermail/discuss/attachments/20210826/d26a5385/attachment.png>