Beyond the Buzz: Expert Panel on AI & the Modern Firm

Start by identifying which tools you already have access to: Microsoft 365, Google Workspace, QuickBooks, or others. Then define the specific problem you're trying to solve—don't jump to AI first. Look for the work that's eating up the most of your time: if you're spending six hours weekly on a repeatable task, investing a few hours to learn an AI tool makes sense; if it's thirty minutes, probably not. Once you've picked your problem, pilot something basic with your team before scaling it up.

You likely already have AI available to you. If you're using Microsoft 365, Copilot is built in; if you're on Intuit, QuickBooks, or Xero, those platforms are embedding AI capabilities. You can also access Claude or ChatGPT models through Copilot's premium tier, so you don't necessarily need separate subscriptions. The key is understanding what your current systems offer and whether the tier or security settings are configured correctly for your needs.

You'll be able to assess whether your firm should be watching, experimenting, piloting, or actively running AI—and understand what leadership decisions come next. You'll know how to use AI safely with client data: for example, cleaning up messy price lists or charts of accounts in Excel, or catching anomalies in financial data while staying compliant. You'll also understand the trade-offs between different tools (Copilot, Claude, ChatGPT) and how to build an AI policy for your firm that protects both you and your clients.

This course is grounded in practical workflows and real firm scenarios, not marketing promises. The panel walks through specific examples: using Copilot in Excel to standardize charts of accounts, the MCP protocol for connecting your AI to your calendar and email to save time, and how to evaluate whether to use Intuit's embedded AI or a standalone tool based on your firm's actual risk tolerance. You'll hear what works, what doesn't, and the trade-offs—including failed AI projects and why some vendor promises don't match reality.

You need three things in place: compliance checks beforehand (know your provincial and federal guidelines), transparency to your client, and a response plan if something goes wrong. The course covers the distinction between operational work (scheduling, emails) and sensitive financial data—you can safely use AI for lower-risk tasks without involving client data. When you do use client information, the instructor demonstrates how embedded AI in platforms like QuickBooks or Excel with enterprise security settings is lower-risk than uploading data to a separate tool, and recommends always asking: what's the risk for what you get out of it?