AI for Business Analysts in Europe: Governance, GDPR, and Practical Use Cases
Key Takeaways
- AI has been part of business analysis for years, but LLMs significantly raise both opportunity and risk
- In Europe, the GDPR makes tool selection and configuration non-negotiable
- Compliance isn’t automatic; it depends on proper enterprise setup and governance
- “Shadow AI” is a greater risk than AI itself when organizations fail to provide safe tools
- Structured experimentation enables safe adoption and builds practical AI expertise
- Standardization of AI tools prevents team misalignment and productivity gaps
- High-value use cases include test case generation, acceptance criteria drafting, learning support, translation, and team facilitation
- AI amplifies judgment and curiosity, but it doesn’t replace business analysis
Disclaimer: The views and opinions expressed in this article are those of the author and may not reflect the perspectives of IIBA.
This article is part of AI Wednesdays, an ongoing 2026 series that explores how business analysis professionals are using AI in real, practical ways. Each article is written by a practitioner and shares experience-based insights to help you use AI with greater confidence—starting small, building familiarity over time, and applying it where it adds real value.
Business analysis professionals often talk about artificial intelligence as if it suddenly burst into our professional lives in 2023 in a blaze of neural network glory. But if we’re honest, we’ve been living with little forms of AI for years.
Take Microsoft Word’s autocorrect. It has always behaved like a tiny, overconfident intern who tries to help but occasionally produces something spectacularly wrong. Recently, I typed “regulatory landscape,” and Word enthusiastically corrected it to “regularly landscaped,” making it sound like I spend my days trimming hedges instead of analyzing requirements.
The point? Even in these small moments, AI has been nudging our workflow. And we’ve already been exercising judgment, critical thinking, and context interpretation long before the rise of Large Language Models (LLMs).
But now that LLMs are maturing and finding their way into regulated industries, the impact on our work as business analysis professionals is far more significant. So are the opportunities (and the risks).
Working in the highly regulated European banking sector, I’ve been observing how AI changes our role, our methods, and even our collaborative habits. This article shares those field insights. From compliance considerations to experimentation, from frustrations to genuinely useful use cases.
European Regulation, GDPR, and the Rise of “Shadow AI”
Let’s get the serious part out of the way first.
Business analysis professionals working in Europe are always faced with three realities: regulation, compliance, and documentation. And whenever AI enters our workflows, those realities grow sharper.
1. The Non-Negotiable: GDPR
The General Data Protection Regulation (GDPR) imposes strict rules on how personal data is handled. For AI tools, this includes:
- Where the data is processed
- How the model uses (or doesn’t use) the data we submit
- Whether prompts might leak sensitive information
- Whether the service meets organizational privacy expectations
This alone makes the choice of LLM crucial.
2. Picking the Right Tools
In my own practice, I use Mistral’s “Le Chat” precisely because it’s natively compliant with European standards and built with privacy by design. It gives me peace of mind without requiring acrobatics to avoid breaching confidentiality.
In parallel, I also work with organizations using the enterprise version of Microsoft Copilot. This version can be compliant, but only when the organization activates and configures the right safeguards.
This is an important nuance for business analysis professionals: compliance isn’t a feature but rather a configuration.
3. The Real Threat: Shadow AI
The biggest risk I see today isn’t necessarily the technology itself, but what is called “shadow AI”: business analysis professionals using AI tools outside the organizational boundaries.
It rarely comes from bad intentions. Usually, it’s because:
- The organization hasn’t yet provided a safe AI tool
- Curiosity drives us
- Deadlines push us
- We want to learn
- Honestly, it’s just too tempting to try new things
But shadow AI can quickly lead to accidental data exposure, undocumented decisions, or inconsistent practices.
For organizations, the lesson is simple: if you want business analysis professionals to avoid risky tools, give them safe ones. And, more importantly, give them a safe environment to learn, test, explore, and experiment.
Empowering Business Analysts to Experiment with AI
Once the right tool is selected, it’s time to free the business analysis professional. Not completely, mind you, but enough to allow creativity and exploration. The organizations that succeed with AI are the ones that create structured freedom, and my preferred approach follows four steps.
Step 1: Exploratory Testing — The Joy of Playing
Let business analysis professionals interact freely with the AI. This first phase is about curiosity, intuition, judgment, and the courage to try things even when you don’t know if they’ll work.
The only guideline is to start with a very small scope. Something easy, low risk, and quick to evaluate. Here are some examples of what I mean:
- Rewriting a user story
- Generating acceptance criteria
- Turning a meeting summary into action points
- Asking for a conceptual explanation of a regulatory concept
If the experiment fails, you shrug and move on. If it works, it’s a small victory that sparks confidence.
Step 2: Share Findings (and Feelings)
This is where the magic happens. Try to share whatever you learn from your AI experiments with your fellow business analysis professionals, and ask them for the same:
- What worked well?
- What didn’t?
- What surprised them?
- What frustrated them?
You’ll quickly see patterns emerging. Similar use cases raise enthusiasm; similar pitfalls create friction.
From this collective intelligence, a shortlist will gradually develop, containing the low hanging fruits: tasks where AI is fast, consistent, reliable, and genuinely helpful.
Step 3: Explore the Valuable Few
Once you’ve identified the top use cases, focus on them. No need to automate everything. No need to industrialize chaos.
Pick the two or three use cases with the biggest return on effort. Then refine prompts, compare outputs, define good practices, and repeat and improve. This is where business analysis professionals start building AI expertise, shaped by their operational reality, not by generic advice.
Step 4: Stay Curious and Keep Experimenting
AI evolves weekly. What was impossible six months ago is trivial today, so continuous experimentation is key. Not in a chaotic way, but as a set of small, structured habits:
- Try one new feature per sprint
- Explore one new prompt technique per month
- Test a new AI-supported workflow every quarter
AI rewards curiosity, and as business analysis professionals, curiosity is already one of our core strengths.
Possible Pain Points
Let’s be honest: implementing AI inside an organization isn’t always smooth sailing. Here are two of the pain points I encounter most often.
1. Unequal Access Creates Frustration
Some business analysis professionals get full access to a powerful LLM, while others get nothing. This unequal playground naturally creates friction: “Why can they automate test cases while I still do everything manually?”
It’s human. It’s predictable. And organizations should anticipate it.
2. Different Versions Create Misalignment
Even when everyone does have access to an LLM, they often don’t have the same one. A common example is:
- Some analysts have Copilot Web (the light version)
- Others have Copilot with a licence (the full version)
Same brand, but completely different capabilities. This leads to confusion during collaboration. One person can run advanced features, the other can’t, and the team loses time aligning outputs.
The lesson? Standardization matters, even for AI tools.
My Favourite Ways to Use AI as a Business Analyst
After months of experimentation, here are the use cases I personally find to be the most valuable.
- Generating test cases and Gherkin acceptance criteria: This is probably the most effective use case for me. You provide a user story, a workflow, or a business rule, and the AI generates structured scenarios or Gherkin outlines. It’s not perfect, but it’s fast, consistent, and surprisingly useful as a starting point.
- Continuous learning made easy: When exploring a new topic, I often use AI to explain a concept in simple terms, get a high-level summary of a complex domain, compare two regulatory approaches, or maintain my “technology watch.” It’s like having a knowledgeable colleague who is available 24/7 and eager to help.
- Image generation for team rituals: One of the most fun discoveries has been using AI to generate playful images as icebreakers for sprint reviews or retrospectives. Before AI, I had the artistic ability of a potato. Now, I can create a cyberpunk elephant skiing. These small touches inject some much-needed humour and energy into team rituals.
- Translation and writing assistance: I often write drafts in French (my mother tongue), then use AI to translate or polish them into English. This article, for example, has been polished with the help of AI.
AI Is a Companion, Not a Replacement
AI will not replace business analysis. Business analysis professionals who know how to use AI will certainly outperform those who don’t. The real value for us lies not in automation, but in amplifying our curiosity, our strategic thinking, our ability to analyze faster, and ultimately our impact.
In the European banking sector (or in any regulated environment), the winning approach is simple:
- Choose compliant tools aligned with regulation
- Empower business analysis professionals to experiment safely
- Share knowledge and build collective intelligence
- Focus on use cases that genuinely bring value
- Keep improving, exploring, and thinking critically
AI isn’t the destination. It’s simply a tool that helps us navigate complexity with sharper judgment, greater creativity, and a little more humour.
Join IIBA’s global community to exchange AI use cases, governance insights, and practical lessons from the field.
About the Author
Stéphane Depauw is a certified project manager and business analyst with over 20 years of experience, specializing in finance and digital transformation. As the founder of Appetite For Solution, he helps organizations turn challenges into practical solutions by focusing on their real needs and operational realities. His work blends hands-on expertise with a passion for entrepreneurship and innovation. Beyond consulting, Stéphane draws inspiration from pop culture, economics, and the unexpected connections between them, always looking for fresh perspectives to solve complex problems.
