As long as we never KYC, I guess they'll not interfere so much on a surveillance level. Although A.I. models are a looming threat in that respect, for malicious actors (the States, the GAFAM corporations,...) to forecast our decisions.