We propose a Copilot- or AI-driven analysis feature that flags financial transactions which significantly deviate from historical posting behavior.

• Learn “normal behavior” based on:
• GL account usage
• Posting patterns (amounts, frequency, dimensions)
• Time of year or period
• Automatically highlight or tag:
• Unexpected high/low amounts
• Rare GL account usage
• Missing expected dimension values
• Display alerts in journals, ledger entries, or as part of a review dashboard