Privacy — Social Networks and Web Tracking
Stub. What privacy means as a security-adjacent property distinct from confidentiality; social network privacy failures (over-sharing, re-identification from "anonymized" social graphs, inference from friend networks); and web tracking mechanisms (cookies, fingerprinting) and the countermeasures/regulation built against them.
Privacy is a distinct property from confidentiality: even data that’s never technically leaked can violate privacy if it enables unwanted inference. Social networks are the sharpest example — individually innocuous posts, likes, and connections compose into detailed inferences about a person that outstrip what they intended to share, and even graphs “anonymized” by stripping names have repeatedly been re-identified using structural properties alone. Web tracking is the same problem applied to browsing behavior: cookies (first- and third-party) and fingerprinting (using enough incidental browser/device characteristics to re-identify a user without cookies at all) let advertisers and trackers build profiles users never explicitly agreed to, motivating both browser-level countermeasures and legal/regulatory responses.
Outline (TODO — flesh out each)
- Privacy vs. confidentiality
- Social network privacy failures — over-sharing, re-identification
- Inference from social graphs
- Cookies — first-party vs. third-party tracking
- Browser/device fingerprinting
- Countermeasures and regulation