Every L&D and HR leader we speak to has a version of the same story. People are not engaging with learning programmes, and if they do even launch or sign up then completion rates for non- mandatory learning is less than 20%. The instinctive response is to ask what is wrong with the content, or with employees. The more useful question, is what is stopping people from learning, given that upskilling is the top criteria for staying employed and relevant.
Disengagement is being misread
Low engagement with workplace learning is almost universally described as an apathy problem. Employees are not motivated, attention spans have shortened, and people are too busy to develop. These are reasonable-sounding diagnoses, but they point to the wrong place, and the strategies they produce tend to make things worse rather than better.
What we are observing in most organisations is not apathy. It is overwhelm. Employees are navigating a significant volume of cognitive and emotional demand: content libraries with hundreds of courses and no clear starting point, an expectation of continuous reskilling on top of existing workloads, and for many people, genuine uncertainty about which skills are worth investing in and what their career will look like in a few years. When that level of uncertainty and information volume collides, disengaging is a rational response, not a motivational failure. People are not opting out of learning because they do not care about their development. They are opting out because the conditions for development have not been created.
Why the attention span argument does not hold up
The idea that modern employees have an attention span of around eight seconds, roughly equivalent to a goldfish, became widely cited in L&D and was used to justify shorter, snappier content design. The underlying research does not support it as a meaningful guide to how people learn at work.
Attention is selective, not fixed. People concentrate on things that feel relevant to them, where they understand the purpose, and where they have the cognitive capacity to engage. Remove any one of those conditions and attention fails, not because it has diminished, but because it was never given a reason to mobilise. The question worth asking is not how to make content shorter, but why the content does not feel relevant or manageable to the people it is intended for. Those are design and strategy problems, not physiological ones.
Career anxiety as a learning blocker
One factor that is consistently underestimated in L&D strategy is the role of career anxiety in blocking learning before it even begins. Research from Mental Health UK (2025) and Microsoft’s Work Trend Index (2025) points to employees experiencing substantial uncertainty about their professional futures, and AI is accelerating this not simply by threatening specific roles, but by making it genuinely less clear what good career development looks like in the first place. When someone does not know what they are developing towards, asking them to engage with a learning programme is asking them to invest time and energy in something that feels directionless. Anxiety does not just affect confidence. It occupies cognitive bandwidth in ways that make it harder to concentrate, to take in new information, and to feel that any of it will make a difference.
Gallup’s data on UK employee engagement consistently shows that employees who feel unclear about their own development are significantly less likely to engage with workplace learning, not because they do not want to grow, but because growth feels inaccessible when direction is absent.
What learners say they actually need
Employees are not asking for more content. What they want is confidence that the time they invest will lead somewhere meaningful. That is a significant reframe, because it shifts the question from how much we are offering to whether people leave learning with more clarity and confidence than they arrived with.
The research identifies three things learners consistently prioritise above everything else: clarity about how learning connects to their actual future, relevance that is obvious and immediate rather than theoretical, and guidance rather than simply access to resources. Flexibility without direction, as the data bears out, just feels like drift. This connects directly to the way bespoke digital learning can change the conditions for engagement, because when learning is built around a person’s real situation rather than a generic competency framework, the experience of it is fundamentally different.
How AI can reduce the cognitive burden that is blocking development
The case for AI in this context is not primarily about efficiency or automation. It is about removing the cognitive burden that is preventing development from happening. Done well, AI can do something that traditional L&D approaches genuinely struggle with: meeting people where they are, helping them understand what they actually need to work on, and clearing enough of the noise so that getting started feels possible rather than overwhelming.
Research from The Conference Board found that AI-enabled development can deliver approximately 90% of what day-to-day career coaching provides, with goal attainment more than doubling when AI support is in place compared to traditional approaches. Those outcomes depend, though, on whether the AI has been designed to reduce cognitive load rather than add to it, because an AI tool that presents fifty personalised recommendations is not addressing overwhelm, it is repackaging it in a more sophisticated form.
The data from our own pilot
At Saffron, we ran a four-month pilot of our AI coaching platform, with 500 users. The engagement metrics were strong: over 50,000 minutes on the platform, 628 development actions completed, and 79% of participants finishing their action plans within the agreed timeframe. But the figures that tell the more important story are the psychological ones.
Confidence in career progression moved from 36% to 92% over the course of the pilot. Sense of control over career development increased from 50% to 65%. Concerns about future development dropped from 37% to 10%, a 73% reduction in development-related anxiety over four months. And 90% of participants discovered a development pathway they had not previously considered. That last figure reflects what happens when you reduce the noise enough: people do not simply engage with learning, they find possibilities that were always available to them but obscured by information overload.
The pilot was for the Department for Education and was independently monitored and evaluated by the Learning and Work Institute. It’s design was approved by Livework Studio – a consumer design strategy consultancy.
The shift L&D strategy needs to make
The implication for how L&D operates is real, even if it requires rethinking some established practice. The dominant model, identifying what people should know, building content, and pushing it out, was not designed for the psychological reality most employees are now living in. The emerging approach is less about content delivery and more about managing cognitive load, less about measuring completions and more about measuring whether people feel more confident and capable as a result, and less about providing access and more about creating conditions where learning feels genuinely possible in the first place. This is the sweet spot where learning can actually start. You still need good relevant content for learning to actually happen.
Your people are not apathetic. They are overwhelmed. That is a meaningfully different problem, and one that L&D, with the right strategy and tools, is very well placed to address.
To find out more about how AIDA supports learner confidence and engagement, visit the AIDA page or get in touch with the Saffron team.
