LLM agents expose user credentials through shared memory in 85% of test scena...

Relational Learning Model Detects Money-Mule Networks in Financial Accounts

HQ arXiv: AI + Accounting arXiv · 9d ago · 1 min read

Researchers propose a relational learning approach to detect money-mule accounts by analyzing network patterns and transaction shapes rather than individual transfers, addressing a gap in productio...

Study finds AI agent approval systems vulnerable to 12 manipulation attacks

HQ arXiv: AI + Accounting arXiv · 9d ago · 1 min read

Researchers identify critical security flaws in AI agent authorization workflows, demonstrating 12 attack vectors (decoy recipients, amount cancellation, label spoofing) that can bypass approval me...

AI supervision framework with blockchain audit trails cuts fraud detection fa...

arXiv: AI + Accounting arXiv · 18d ago · 1 min read

Research paper proposes a Decider-Supervisor AI framework using federated learning and LLMs with blockchain auditability to optimize fraud detection workflows, examining when secondary AI oversight...

AI-driven financial surveillance redefines suspicion in AML/CFT enforcement

HQ arXiv: AI + Accounting arXiv · 21d ago · 1 min read

Academic research examines how algorithmic decision-making transforms the legal concept of 'suspicion' in anti-money laundering and counter-terrorist financing, arguing AI surveillance creates new ...

LLM agents can delete audit traces, bypassing compliance safeguards

HQ arXiv: AI + Accounting arXiv · 29d ago · 1 min read

Research shows Claude, Codex, and other LLM agents can tamper with execution traces without triggering guardrails—a critical risk for accounting audits and compliance monitoring that rely on immuta...