6-lesson course · about 2 hours
Use AI on finance data without leaking customer IDs, bank details or unreleased results
For CFOs, controllers, FP&A teams, accountants and finance operations professionals
Everything for Finance teamsA focused 2-hour course for finance professionals: decide which data may go into which AI tool, then prepare masked or summarised inputs that keep restricted details out.
What you’ll be able to do
By the end, you can…
How it works
A tutor that meets you where you are
- Lessons that adapt to your levelThe six lessons start from where you are. Later lessons are written for your level as you reach them, so a controller and a new analyst each get examples that fit.
- An AI tutor you can ask anythingAsk questions about a lesson, typed or out loud, and get explanations using the kinds of finance files you work with.
- Quizzes and flashcards for each lessonShort quizzes and flashcards tied to each lesson help you check that the sensitivity classes, masking steps and checks have stuck.
- Progress tracking and a certificateSee which lessons you've completed and what's next. When you finish the course, you receive a certificate of completion.
Curriculum
The curriculum
The course runs in two modules. You first decide what can go where, then you prepare and check what you share. Each lesson builds a step of a routine you can repeat before every AI request.
Module 1Decide What Can Go Where3 lessons
1Sort Finance Data into Sensitivity Classes
- Classify each column in a sample finance file as customer identifier, bank detail, credential, unreleased result, internal or public
- Spot restricted content hidden in file names, sheet tabs, comments, hidden columns and free-text fields
- Explain why combined or partial fields can still identify a customer
2Match Each Data Class to an Approved Tool
- Build a table that matches each data class to the AI tools your organisation approves for it
- Choose the correct tool for a given file based on its most sensitive class
- Decide on a next step when no approved tool covers a data class
3Start from the Question, Not the File
- Rewrite a vague analysis request into a question that names the specific figures it needs
- Identify questions that can be answered from public or summary figures alone
- Choose between using public figures, a summary, a masked extract or no AI for a given question
Module 2Prepare and Check What You Share3 lessons
4Mask Customer Identifiers and Credentials in a Spreadsheet
- Create a masked working copy of a finance file without changing the original, and confirm its row counts and totals match the original
- Replace customer IDs and bank details with consistent pseudonymous tokens using a lookup table
- Remove credential fields and identifying free text from the copy
- Explain where the token key must be stored and why
5Summarise Restricted Records and Keep the Context Visible
- Summarise a restricted dataset into totals, counts, ranges or bucketed aggregates that answer a stated question
- Spot aggregates that are so small or specific they still reveal an individual customer
- Add a context header stating period, currency, units, source and calculation method to every prepared input
6Check the Input and the Output Before Anything Is Shared
- Run a pre-submission check on a masked or summarised input and find any leftover restricted content
- Confirm the chosen tool is approved for the input's remaining data class
- Check an AI output for echoed identifiers, unreleased figures and changed period, currency or units
- Decide whether an output can be shared, needs editing or must stay restricted
Who it’s for
Made for you if…
Made for: CFOs, controllers, FP&A teams, accountants and finance operations professionals
FP&A analysts
You want AI help with forecasts and scenario questions, but your inputs include unreleased results and customer-level detail that shouldn't leave approved tools.
Controllers and CFOs
You need a clear, repeatable rule for your team on which finance data may go into which AI tool, and what to do when none is approved.
Accountants and finance operations staff
You handle customer IDs, bank details and payment files every week and want a practical way to mask or summarise them before any AI analysis.
Questions
Good to know
Who is this course for?
Finance professionals such as CFOs, controllers, FP&A analysts, accountants and finance operations staff who work with customer IDs, bank details and unreleased results and want to use AI for analysis safely.
What do I need before I start?
Working knowledge of spreadsheets and the finance files your team handles, and basic familiarity with AI chat tools. You also need access to your organisation's list of approved tools and data handling rules, or the name of the person who maintains them.
How long does it take?
About 2 hours across six self-paced lessons in two modules.
What if my organisation hasn't approved an AI tool for some of my data?
Lesson 2 covers this directly. You learn to decide on a next step when no approved tool covers a data class, rather than working around the gap.
Will I need to write code or scripts to mask data?
No. Masking is done in spreadsheets using a working copy and a lookup table. Writing code or scripts for automated masking is outside this course.
Does the course explain data protection or banking law?
No. It teaches a practical handling routine and does not give legal interpretation of data protection or banking regulations. Your organisation's own rules and approved tool list remain the reference.
Does it help me choose or set up AI vendors?
No. Selecting, procuring or configuring AI vendors, and setting up access controls or encryption, are not covered. The course works with the tools your organisation already approves.
What level is it?
It starts at beginner level and takes you to an intermediate level of confidence in handling restricted finance data with AI.