One could say that the idea for device-independent security of quantum key distribution goes back to Ekert’s famous E’91 paper, which is well worth a read (and it is short!).

The word “device-independent” was first coined much later, in a paper by Acin et al. Maybe the first formal argument is due to Mayers and Yao, who argued that the CHSH correlations could be used to “self-test” the EPR pair.

A lot of work has happened in device-independence since then. A good resource for more bacgground information is this recent editorial for a focus issue of the New Journal of Physics devoted to device independence. You may also wish to consult this extensive survey by Brunner et al. on “Bell Nonlocality”.

In Week 10 of the course we will return to some of the ideas expressed this week, to show that the CHSH game and its “rigidity/self-testing” properties can be used to obtain much more than device-independent quantum key distribution: delegated computation, or “a classical leash for a quantum system“!