A recent UCL research reveals AI video can match human instruction in impact and engagement.

Video has long been a pillar of workplace training. But a fundamental question remains: Can AI-generated video — created without cameras, actors, or studios — really rival traditional instructional formats?

A new study from University College London (UCL) offers a compelling answer: yes. 

And for L&D professionals seeking speed, scalability, and engagement without compromising learning outcomes, that answer couldn’t come at a better time.

 

The Research: Putting AI-Generated Video to the Test

Led by PhD researcher Zoe Ruo-Yu Li at UCL’s Knowledge Lab under the supervision of Professor Mutlu Cukurova, the study examined the impact of AI-generated video on adult learning.

It compared four different formats — AI-generated video with an AI avatar, human-recorded video, AI-written text, and human-written text — to evaluate how each affected learners’ recall, recognition, and emotional response.

The study involved 500 adult learners and used a mixed-methods, pre- and post-test design to measure both cognitive outcomes and affective feedback. The learning topic was workplace food safety — relevant yet unfamiliar for most participants.

 

Finding #1: AI-Generated Video Matches Human-Recorded Video in Perception and Learning Outcomes

The study revealed no statistically significant difference in learning performance between human-recorded and AI-generated videos. Both formats enabled learners to recall and recognise information with similar success rates.

In other words, digital avatars can be just as effective as human presenters — as long as the instructional content is sound.

This levels the playing field for AI video in corporate training environments. It’s no longer a novelty or fallback option — it’s a credible learning tool.

 

Finding #2: Video Beats Text in Engagement and Preference

While performance was relatively equal between the two video formats, video — in any form — clearly outperformed text.

Key data points:

• 77% of learners preferred video over reading text.
• 94% wanted more video-based training content in their workplace.
• Emotional engagement was significantly higher with video than text.

This confirms a growing trend: learners are increasingly drawn to rich, visual formats — and they expect modern training to reflect that.

 

Finding #3: Faster Completion Time — Without Sacrificing Learning

Interestingly, the study found that participants who watched the AI-generated video completed the training significantly faster than those who watched the human-recorded version.

On average, viewers spent 20% less time on the course when it was delivered via synthetic video — thanks to its more concise format. Crucially, this reduction in time had no negative impact on learning outcomes. Recall and recognition scores remained just as strong.

For L&D teams under pressure to deliver efficient learning at scale, this signals a powerful opportunity: AI video can save valuable learner time without compromising effectiveness.

What This Means for L&D Professionals

AI video isn’t just about novelty — it’s about solving real operational problems in training:
– Cost
– Speed
– Scalability
– Localisation
– Content maintenance

But a word of caution: not all AI video platforms are created equal.

 

Choose Tools Built for Learning

Many AI video generators are designed for marketing or pure entertainment. But platforms built specifically for workplace learning — like Synthesia — are different by design.

They understand:

✔️ How L&D teams structure and scale content.
✔️ What compliance and data privacy standards apply.
✔️ The importance of version control, localisation, and rapid updates.

Synthesia comes with purpose-built workflows that reflect how instructional designers and enablement teams actually work. That means less friction, fewer workarounds, and more time saved across the board.

 

Here’s what that looks like in practice:

Faster production cycles

Go from script to video in minutes — no cameras, studios, or editing tools required.

Team collaboration

Work together across teams and regions with shared access, edit history, comment threads, and approval workflows built into the platform.

Scalable localisation

Create videos in 140+ languages using the same avatar and script, without reshooting or revoicing.

Consistent delivery

Ensure message accuracy with AI avatars that always stay on script, no matter the language — ideal for compliance and policy training.

Easy iteration

Update a single line of dialogue or entire scripts instantly — regenerate videos without starting from scratch.

Analytics that close the loop

Track engagement, completion rates, and drop-off points — then improve your content based on real data.

Built-in interactivity

Add clickable elements, quizzes, or branching scenarios to transform passive viewing into active learning, all within one platform.

Multilingual video player

Automatically detect the viewer’s browser language and let them switch between language tracks on demand — no duplicate video files required.

Enterprise-grade safety

Choose a platform built with AI ethics and safety in mind — from robust content moderation to responsible development practices that support secure, scalable learning.

 

Final Thoughts: A Turning Point for L&D

The UCL study is more than just an academic milestone — it’s a signal that AI-generated video has earned its place in the L&D toolkit.

It proves what many forward-thinking teams already suspected: when done right, synthetic video can deliver meaningful learning outcomes, engage learners, and scale effortlessly across global organisations.

As AI video tools continue to evolve, the gap between human and synthetic delivery will only shrink — and in many cases, synthetic will be the faster, safer, and more adaptable option.

For L&D leaders, the question is no longer “Is AI video good enough?”
It’s “How can we use it to design learning that’s future-ready?”

The opportunity is here. The research backs it. Now is the time to build smarter, faster, and more scalable learning experiences — with AI video at the centre.

 

Ready to try it for yourself?

Synthesia is trusted by 70% of Fortune 100 companies and is built specifically for workplace learning — from rapid video creation to multilingual delivery, team collaboration, interactivity, and analytics. 

You can get started for free and see how easy it is to bring your training content to life with AI video.

 

Alexandru Voica 

Head of Corporate Affairs and Policy at Synthesia