Reconstructive Identity: LLMs Link Weak Signals to Deanonymize at 68% Recall
AmuzedX · reddit · 2026-09-28
A Reddit long-form paper abstract proposes the concept of reconstructive identity: identity need not be stored as an explicit field—it can be inferred by correlating weak signals scattered across text, images, behavior, metadata, and public records.
Key evidence
- Cites Lermen, Paleka, Swanson, Aerni, Carlini, and Tramèr's study: LLM-based cross-platform deanonymization achieves up to 68% recall at 90% precision, far above classical baselines
- Stylometry (including source-code style) and behavioral biometrics (gait, voice, hand movement, gaze) already function as persistent identity signals
- Computer-vision research increasingly treats natural-language descriptions and visual observations as interoperable person re-identification signals
Core argument
The unit of privacy risk is no longer the explicit identifier but the linkability of ordinary information. Individual defenses have limits; future privacy protection must address what identities systems can reconstruct, not just what they store.
More from Safety
- Medicare 'breach' may not be a breach — the real story is how OpenAI's agent telemetry caught it — taotau · 2026-09-28
- GPT-6 Astra system card: CoT monitor recall drops below 11%, latent reasoning kills monitorability — enginetown · 2026-09-28
- VPNs don't hide your location: timezones, WebRTC and DNS leaks give you away — StewartalsopIII · 2026-09-28
- OpenAI agents hit UN trade database 16,000+ times, bypassing anti-bot filter — CtrlAltDwayne · 2026-09-28
- AI-Powered Phishing Scam Nearly Hijacks a 10-Year-Old X Account — StewartalsopIII · 2026-09-28
- Doomers vs cybersecurity: control is possible, but frontier labs show no will to do it — trevposts · 2026-09-28