From ADDIE to ADGIE: A learning design framework reimagined for the AI era
Aug 27, 2025
7
min
In the world of corporate learning, certain acronyms have stood the test of time. One of them is ADDIE — Analyze, Design, Develop, Implement, Evaluate — long considered the gold standard for instructional designers. But in an era shaped by ChatGPT, DALL·E, and other generative AI tools, this linear framework is starting to show its age. While doing some summer L&D research, I came across a great academic paper by researchers Khadija Hilali and Meriyem Chergui. In it, they propose a bold rethink of the ADDIE model — one that integrates artificial intelligence from the ground up. Enter ADGIE, a new instructional design model that weaves AI into every stage of the process without ever sidelining the human designer. Here’s why I believe this framework could shape the future of augmented learning design.
©Blify
Why ADDIE no longer works for today’s learning needs
Developed in the 1970s to guide the creation of structured training programs, ADDIE is still widely used in businesses, universities, and training organizations around the world. It’s structured, adaptable, and time-tested.
But the workplace has changed — fast. And learning technologies have exploded. As a result, ADDIE shows its limitations:
Long production cycles clash with the need for speed and agility.
Limited personalization fails to engage today’s diverse learners.
A lack of native integration with modern digital tools, especially AI.
As Hilali and Chergui argue, “Traditional models like ADDIE lack mechanisms for personalization, human-machine collaboration, and iterative validation loops.” The result? Learning experiences that are often rigid, misaligned with learner needs, and slow to evolve in fast-moving work environments.
Meet ADGIE: a framework built for the AI age
ADGIE — short for Analysis, Design, Generation, Individualization, Evaluation — isn’t just ADDIE with a chatbot glued on. It fundamentally redefines who does what in the instructional design process. Every step is mapped to either the human designer, the AI assistant, or both — depending on their respective strengths.
The model is built on 3 key principles:
Explicit human-AI collaboration, made visible via role assignments at each stage.
Iterative cycles, with continuous validation from humans and learners alike.
Advanced personalization, powered by real-time learning data.
Put simply, ADGIE is ADDIE supercharged by AI — but grounded in human oversight.

A closer look at the 5 stages of the ADGIE model
1. Analysis
AI gets involved from the start, using natural language processing to generate learner personas from survey responses or qualitative feedback. The designer then reviews and refines the insights.
2. Design
Humans still lead on learning strategy. But AI helps structure content, suggest learning pathways, and propose scenario templates — dramatically reducing production time without sacrificing quality.
3. Generation
This is where AI shines: creating videos, visuals, audio clips, quizzes, and more. But the designer remains the conductor — reviewing, tweaking, or rebuilding content to ensure alignment and impact.
4. Individualization
Thanks to real-time learning data, AI adjusts content based on each learner’s preferences, gaps, or engagement patterns. A video instead of a text explanation. A deeper dive into a tough topic. The learning path becomes adaptive, dynamic, and personalized.
5. Evaluation
This is the model’s beating heart — a continuous feedback loop involving both learners and designers. The aim: maintain instructional coherence, spot AI blind spots, and improve resources over time.
Webinar
How to Scale and Embed Your Managerial Culture with AI: Real-world insights from Club Med, market trends, and a live demo of the right tool for the job
What the research tells us: L&D professionals are ready
To develop ADGIE, the authors surveyed 90 education and training professionals. The findings show clear excitement for what AI can bring — but also a firm belief in preserving pedagogical rigor.
94% welcome content generation automation.
86% want help with structuring learning content.
91% say they must personally validate AI-generated content.
Only 49% feel “very ready” to delegate pedagogical method selection to AI.
The takeaway: yes to smart automation, no to blind delegation. ADGIE embodies this balance — enhancing efficiency while reinforcing instructional integrity.
Clear gains, real challenges
In my view, adopting ADGIE means embracing:
Faster design cycles, thanks to automated workflows.
Highly personalized learning, tailored in real time.
Elevated roles for learning designers, who focus on coherence, strategy, and innovation.
Stronger alignment with business needs, as learning adapts at the pace of change.
But it’s not a plug-and-play solution. Success with ADGIE depends on:
High-quality data to fuel accurate AI outputs.
Training for instructional designers on AI collaboration.
Ethical guardrails to address bias, access, and data privacy.
ADGIE: a real answer to enterprise learning demands?
Faced with talent shortages, budget constraints, and upskilling pressure, many L&D teams are looking for ways to do more with less. ADGIE might be the framework they’ve been waiting for.
Companies adopting this model could:
Scale instructional design without compromising quality.
Roll out adaptive content quickly across regions and teams.
Elevate the trainer’s role — from content builder to experience curator and human connector.
Some early adopters, like Safran and LVMH, are already experimenting with generative tools like Synthesia and GPT Creator alongside traditional ADDIE workflows. ADGIE takes it a step further by integrating AI from the very start of the process.
Conclusion: a compass for the next generation of learning designers
ADGIE isn’t a buzzword. It’s a thoughtful, grounded attempt to build a learning design model that reflects both the power of AI and the responsibility of humans. It doesn’t replace instructional designers — it empowers them.
It lays the groundwork for agile, ethical, learner-centered training at scale. It still needs real-world validation in corporate and academic settings, including vendors like Blify. But the foundation is solid.
One thing is certain: as companies scramble to “do more with less,” ADGIE shows a promising path forward — where AI is an enabler, not a replacement. Where the designer stays at the helm, more equipped than ever.
⸻
FAQ
What is the ADGIE model?
ADGIE is an evolution of the ADDIE instructional design model that integrates artificial intelligence into every step of the learning design process — while preserving human oversight.
How is ADGIE different from ADDIE?
Unlike ADDIE, ADGIE introduces a generation phase (AI-powered content creation) and an individualization phase (real-time personalization for each learner).
Why adopt ADGIE in corporate training?
ADGIE helps reduce design time, tailor learning experiences, and integrate AI — all without compromising instructional quality.
Will AI replace instructional designers?
No. ADGIE is built on collaboration between AI and humans. AI handles repetitive tasks, while designers lead strategy, validation, and creativity.
Is ADGIE being used already?
ADGIE is a recent framework still being validated. Organizations like Blify are exploring its application in real-world learning environments.
Author(s)

Clément Lhommeau
Cofounder, Blify
Share article