Source for cryptographical randomness of nRF52-series DeviceID register values

[Otso Jousimaa]

Hello,

The DeviceID register of nRF52 series is referred as "unique", "random" and some answers from staff on this forum claim that the Device ID is generated with FIPS-approved RNG. 

We're using DeviceID as unique ID of devices, and with the updates to CE RED cybersecurity regulation we'd need to present some evidence that our device IDs are actually cryptographically random. Is there some data sheet / reference manual / other official document that explicitly states that DeviceIDs are "best practise cryptographically random" or something among those lines? If not, can I email a Nordic Semi representative for a written confirmation? 

It would help us a lot in dotting our i's and crossing our t's.  

[tesc]

Hi,

We do not have any public datasheet, reference manual or other official document that explicitly states that DeviceIDs are "best practice cryptographically random" or similar, no, if what has been stated by other Nordic employees in other DevZone tickets is not enough.

If you need more than that, then let me know and I can see what else we can potentially state on the matter.

Regards,
Terje

[Otso Jousimaa]

Thank you for the quick reply. I'll mark it as "confirmed by Nordic Semi employees at dev zone + link" for now and reach out to our contact person if we need more formal confirmation.