Towards Better Privacy Reasoning
Bitcoin is often described as private but in reality it is pseudonymous not anonymous. Leaking even small pieces of information can expose a user’s identity or enable long term tracking. Yet privacy is rarely broken by a single catastrophic failure. More often it degrades gradually as information from different sources is combined over time. This talk explores how modern chain analysis works from clustering heuristics and wallet fingerprinting to intersection attacks and metadata leakage and examines how privacy assumptions can break down in subtle and unexpected ways. Even strong guarantees made within one model may fail to generalize once additional information becomes available to an adversary. Rather than treating privacy as a binary property the talk frames it as a spectrum shaped by assumptions, adversaries, and observable leaks that compound over time. Drawing parallels to side channel attacks, it argues that small leaks can accumulate into significant privacy failures even when individual components appear secure in isolation. The goal is not to argue that privacy is impossible, nor to present a single “magic bullet” solution. Instead the talk encourages more rigorous and holistic thinking about how privacy systems behave in the real world and why small improvements across wallets, protocols, infrastructure and user behavior can meaningfully strengthen privacy over time.