Researchers develop LoRA-MINT to audit training data in fine-tuned LLMs

arXiv: AI + Accounting · Gonzalo Mancera, Daniel DeAlcala & 4 others · June 5, 2026 · 1 min read · Relevance: 62/100

New methodology LoRA-MINT enables auditing of training data membership in domain-adapted LLMs, addressing IP and data sensitivity concerns for accounting applications.

Researchers propose auditor-assisted verification for LLM identity in hosted ...
Researchers propose reusable ledger system for auditing LLM agent compliance
Researchers develop auditing method to detect training data leaks in fine-tun...

Get daily agentic AI accounting news in your inbox

Read original article →

Share: Post Share Email

Researchers propose auditor-assisted verification for LLM identity in hosted ...

HQ arXiv: AI + Accounting arXiv · 3d ago · 1 min read

Academic paper introduces framework for auditors to verify which LLM model actually processed requests in hosted services, addressing opacity in AI model deployment claims without requiring runtime...

Researchers propose reusable ledger system for auditing LLM agent compliance

HQ arXiv: AI + Accounting arXiv · 6d ago · 1 min read

Researchers introduce a framework that records AI agent actions into policy-independent ledgers once, then efficiently audits them against multiple compliance policies without re-reading original t...

Researchers develop auditing method to detect training data leaks in fine-tun...

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

Academic paper proposes inLoRA-MINT framework to audit whether specific training data was used in domain-adapted LLMs, addressing data privacy and transparency risks for regulated industries like a...

Study finds LLM compliance systems ignore regulatory rules, threatening audit...

HQ arXiv: AI + Accounting arXiv · 14d ago · 1 min read

Research across 5 LLMs and 20 regulatory domains reveals compliance verdicts remain unchanged even when governing rules are deleted or negated, suggesting AI audit systems may not actually apply ru...

Study reveals evidence-binding gaps in LLM agent audit trails for banking

HQ arXiv: AI + Accounting arXiv · 16d ago · 1 min read

Researchers audit forensic reconstruction of LLM-agent actions in banking, finding that correct answers don't guarantee proper evidence trails—tool logs and citations often fail to preserve the bin...