An HR manager learned that the company’s eLearning authoring tool could now use AI to build a full company-wide course. He had no instructional design background, but he fed it the employee handbook and a set of company policies and produced the company’s new annual anti-discrimination training.
View the original AI-generated course here.
The result was generic content that didn’t reflect what learners actually needed. The knowledge check, interactions, menus, and AI-generated images made the course harder to access and the content harder to apply. Staff were left wondering why they had just clicked through screen after screen of policy and legal language.
Organizations do need employees to understand and follow these requirements. But compliance courses often focus on delivering information rather than on helping people remember it, apply it, and perform. To close the gap between stating expectations and actually changing behavior, compliance training has to be designed for performance. Generating a course is not the same as designing a learning experience.
I redesigned the course to follow a new CSB manager through her workday, guided by these principles:
Performance-based design: Instead of just presenting information, I focused on what learners need to recognize, decide, say, or do when they encounter discrimination or inappropriate behavior at work.
Applying expectations, not reciting them: Learners make realistic decisions and respond to situations, putting expectations into practice instead of repeating rules and legal terms.
Intentional content selection: I included only what employees actually need, rather than pulling in policies and handbook content just because they were available.
Meaningful assessment: Learners show they can apply what they’ve learned in realistic scenarios, not just recall isolated facts.
A cohesive learning experience: The content, activities, interactions, and assessment are designed to work together toward the workplace performance the course is meant to build.
eLearning course built in Storyline 360 supported with Claude Design elements
AI-generated images and audio built in Canva, ChatGPT and ElevenLabs
Job Aid containing key reference information
Interactive features requiring realistic decision-making
Feedback carrying additional learning opportunities
Because the AI-generated course covered so many different topics and was not designed for a specific group of learners, I needed to pare down which information was strictly required for bank employees to know and apply. Creating a scenario-based approach for this eLearning ensured that it would keep learner's attention and also build on their pre-existing background knowledge of workplace behavior.
I chose to have the course follow one character throughout one day to make the learning cohesive and the content approachable. Putting the learner in the driver's seat as they played the role of helper at the various decision points throughout the course means thay they remain active in the learning process.
The knowledge check in the eLearning is scenario-based, therefore designed to assess how well learners can apply the principles into realistic new situations. The course is designed to be delivered together with the job aid, which can be downloaded and saved for future reference.
The results of this eLearning would assessed by looking at the number of CSB discrimination cases filed on a yearly basis (as well as case resolution information), as well as staff and manager surveys.