Regulatory reporting is one of the most challenging roles I’ve worked in - and that’s saying something, considering I worked in public accounting for 10 years, serving large accelerated filers in the audit practice. I know late nights better than I care to admit; add to that the stress of being a cost center, when companies are looking for any excuse to downsize, in a highly regulated world that doesn’t get the funding needed to recruit top talent…it makes for a stressful lifestyle. Where’s the time to learn about AI?
Well, that’s where I can help. I won’t say I’m a master, but I’ve spent the latter half of my career in regulatory reporting, and I use AI tools on real reporting work every day. I started slow, using AI maybe once a week. Over time, I built a strong foundation and embedded it into my daily tasks to drive efficiency and effectiveness for my team.
Most people overthink this. You don’t need a vendor demo, a tech expert, or a steering committee. Whether you’re using Copilot, ChatGPT, or Claude, all you need is a little patience, a desire to learn, and a willingness to try new things. When I give a colleague advice, here’s where I’d start.
Five places to start
Interpret instructions. Build a prompt to help you interpret reporting instructions. Provide the AI with its role, context on its objective, and link it directly to the instructions before asking it to help you research specific reporting questions.
Prepare a leadership presentation. Summarize the results and activities worth highlighting to leadership. Use existing notes, files, and e-mails (public or company-approved data only - see the rules below) to fully empower your AI tool.
Draft the boring writing. Hand it a mock up in bullet points and ask it to draft an executive summary to leadership, an e-mail to your team, or a process narrative to appease internal audit.
Build reporting prep checklists. “Create a checklist for [schedule] from the instructions and process documentation.” Your future self, at 9pm during close, will thank you.
Run a pre-review. “What are the common errors on this schedule, and what should I double-check?” A second set of eyes that never gets tired or takes PTO.
Three rules (the guardrails)
The instructions are public; your numbers are not. Company policy - not the tool - decides what data you can use, and many companies restrict confidential information in AI tools. Confirm what you’re allowed to do before you start.
Verify against source. Treat every answer as a draft. AI will cite a requirement with the confidence of someone who’s never sat through an exam. Treat it like an analyst and review its work.
AI assists; it doesn’t control. It speeds up the draft. It does not sign the workpaper - you still do. Understand the current limitations of AI and minimize your risk of overreliance.
Your first week
Pick one tedious task. Run it through ChatGPT, Claude, or Copilot. Create a prompt to help you either prepare the task, review your results, or summarize key information for wider distribution. Practice using the tools and identifying opportunities for efficiency.
Next issue: What to do when “handy” turns into “I rely on it” - and the controls that have to come along for the ride.
Prepare regulatory reports? Subscribe for the practical, control-aware version of this. And reply with the one task you’d most want AI to take off your plate - I read every reply, and it shapes what I write and build next. Check out some of my free tools!
Educational only; not official guidance or the views of any employer. Always verify against primary-source instructions.