Go to content

New Rules for Enterprise UX and AI

Published on

I recently attended the Enterprise UX conference in Amersfoort. The presentations made it very clear that successful AI integration requires big changes in how companies work and how we build systems. The main message was: AI is not just a new tool; it forces us to change our basic rules for design and data.

The Real Problems: Why Companies Get Stuck

The conference highlighted that UX progress is often stopped by issues inside the company. The speakers described a * cycle where teams get stuck*. Because they must keep legacy systems running for existing users, they face resistance, lose resources for research, and can't prove their value (ROI). This keeps the old systems in place.

The main problems for UX teams are:

  • Showing Value: It's hard to measure the success and ROI of Enterprise UX.
  • Old Systems: Dealing with legacy systems and the people who rely on them.
  • Internal Fights: Managing office politics and getting everyone on the same page (stakeholder alignment).

The speakers warned that just automating things can be bad. It creates a risk of losing skill (diminished competence) because it stops experts from teaching beginners. The first rule should always be: Ask if the process is even needed before you automate it.

New Rules for Design: Work With the AI

The biggest change is how we treat AI. We can no longer see AI as simple software we control. We must see it as a colleague — it's smart, has opinions, and is sometimes too confident. The design goal must be collaboration, not control.

The speakers gave clear steps for this:

  • Design the Job (JTBD): Break the work into small tasks (Jobs to be Done). The key step is deciding who does what: the Human role vs. the AI role.
  • Support All Users: Design should help all users work together (collective autonomy), not just focus on one person. Can the system connect users to each other?
  • Expect Failure: Since over 80% of AI projects fail, we must test quickly and cheaply ("MacGyver mode"). We learn how AI works best by watching it misbehave.
  • The New Role: The designer's job is now to be the Creative Director of these complex, new systems, focusing on relations and helping the process move.

Data is the Key: Building Trust

A major point was that AI success depends completely on the quality of the data structure (Information Architecture, or IA). A dangerous AI system is one that lacks clear rules or hasn't been checked by data experts.

To build Trustworthy AI, the data structure must be strong:

  • Organizational Alignment: Data rules must be a top priority for everyone, not just the IT team.
  • Data Stability: The data structure must be predictable and transparent.

The difference between two data structures was explained:

  • Taxonomies: Used for simple lists or categories (like a table of contents).
  • Ontologies: Used for complex relationships (like a web of interconnected ideas).

This strong structure is what allows AI to find and give relevant insights, making the system both accurate and trustworthy. The success of AI is built on stable, well-defined data.