Insights & resources
Analysis on AI agents for compliance, the AUSTRAC 1 July 2026 reform and onchain financial crime — from the Cogentic team.
The audit trail is the product
Nobody buys software for its audit log. It is the one part that has to work on your worst day.
Where the agent stops
Every compliance vendor with an AI story has to answer one question first. Here is ours, and the line the product is built around.
What is actually in a case note
A summary asks to be trusted. A case note asks to be checked. Most of the engineering goes into the difference.
Seven systems, one decision
If you have worked an alert queue, you will recognise the first half hour of a case. Very little of it is investigation.
Reporter of record: how AI-augmented SARs stay defensible
Procurement teams ask the right question first: who signs the SAR, and what happens to liability when an AI drafts it? Here's the model — and the audit infrastructure that makes it stand up under examiner review.
No rip-and-replace: why agents inside your stack beat case-management migrations
Most compliance vendors ask you to migrate. Cogentic doesn't. The argument for working inside the analytics, KYC, sanctions, custody, case management, and CRM tools your team already runs — and why this is the load-bearing positioning question for the agentic AI category.
The compliance leader's guide to agentic AI in crypto
A working guide for MLROs, BSA officers, and Heads of Compliance evaluating AI agents inside their onchain stack. Validation framework, regulatory mapping (AUSTRAC, MiCA, GENIUS, FATF Rec 16), vendor-selection checklist, deployment playbook.
SOP mapping for crypto compliance: why your procedures need to be encoded
Most VASPs document their compliance procedures in PDFs and spreadsheets. Here is why encoding your SOPs into your compliance platform is the difference between audit-ready and audit-anxious.