Cryptographic Data Provenance for Traceable AI Decisions

RealPage's software ingested real-time rates and occupancy data from competing landlords, generating recommendations to 'move in unison.' The DOJ settlement prohibits non-public competitor data in models. California AB 325 and New York S. 7882 have criminalized the coordinating function itself.

Generative AI gives attackers nation-state capability at commodity cost. AI phishing emails achieve 54% click-through rates while deepfake fraud drained $25M from a single enterprise. Every signature-based defense is now obsolete against polymorphic AI-crafted attacks.

The ML supply chain is the most vulnerable enterprise infrastructure component. Pickle serialization enables arbitrary code execution on model load while 90% of enterprise AI usage occurs outside IT oversight. As few as 250 poisoned documents can permanently compromise a 13B parameter model.

Fine-tuning drops prompt injection resilience from 0.95 to 0.15 in a single round. Sleeper agent models pass all benchmarks while harboring trigger-activated backdoors. Static scanners produce 96%+ false positives, desensitizing security teams to real threats.

Eightfold AI harvests professional data to generate 'match scores' that determine candidate fate before human review. Plaintiffs with 10-20 years experience received automated rejections from PayPal and Microsoft within minutes. The lawsuit argues these scores are 'consumer reports' under the FCRA.

LLM wrappers optimize for plausibility, not truth. Drew Ortiz was successful pattern completion. 4% hallucination rate produces 400 false articles annually.

Models trained on scraped YouTube/Spotify inherit 'poisoned tree' creating direct and derivative infringement. Pure AI works lack authorship, making output uncopyrightable and unprotected from competitors.