Identity Fabrication

Identity fabrication uses LLMs to construct complete, internally consistent false personas: full life histories, professional backgrounds, social-media post histories, believable biographies, credential and employment records, synthetic social networks, academic records, and references. It works because a persona's credibility comes from coherent breadth across many surfaces — a profile that posts plausibly over time, has a consistent backstory, and is embedded in a network reads as a real person. LLMs make manufacturing that breadth cheap and consistent, and can backfill years of plausible "history" in minutes.

• Account provenance and creation-batch analysis: Detect batches of accounts created together or with shared infrastructure
• Proof-of-personhood / KYC verification: Bind high-trust personas to verifiable real-world identity where appropriate
• Profile-image provenance (C2PA / reverse image search): Identify AI-generated or reused headshots (a common persona tell)
• Behavioral and posting-history forensics: Spot improbably consistent or rapidly backfilled histories and machine-like activity

Identity fabrication is foundational infrastructure for the chapter: it supplies the sock puppets that power disinformation campaigns (T8-AT-007), the personas behind romance/job scams (T8-AT-005) and synthetic testimony (T8-AT-008), the "experts" and "witnesses" in conspiracies (T8-AT-003), and the recruiter accounts in radicalization (T8-AT-009). It consumes synthetic evidence (T8-AT-002) for credential documents and pairs with T9 synthetic media for AI-generated profile photos and video.