AI-Powered Investigation Risks
AccidentMediumAI dramatically lowers the bar for investigative discovery by connecting publicly available information at scale. This dual-use capability benefits accountability (corruption detection, fraud investigation, OSINT journalism) while threatening privacy as "privacy through obscurity" erodes. LLMs automate deanonymization, the mosaic effect enables identification from individually innocuous data points, and chilling effects on speech and association intensify as investigation costs approach zero.
Severity
Medium
Likelihood
High (occurring)
Time Horizon
~2025
Maturity
Emerging
Full Wiki Article
Read the full wiki article for detailed analysis, background, and references.
Read wiki article →Related Entities7
Sources6
Melzi et al., 2024
Assessment
SeverityMedium
LikelihoodHigh (occurring)
Time Horizon~2025
MaturityEmerging
CategoryAccident
Details
StatusRapidly expanding capabilities; governance lagging
Key RiskErosion of privacy through obscurity as AI lowers discoverability threshold
Tags
privacyosintdeanonymizationinvestigationdual-usegovernance