Learners today face an overwhelming influx of information. Their attention is fragmented by a steady stream of notifications, emails, meetings and shifting priorities. The result is a growing cognitive load that strains the brain’s natural capacity to process and retain information.

As artificial intelligence (AI) accelerates the pace of change, knowledge workers are expected to learn faster and adapt more quickly or risk being outpaced by AI-savvy peers or even replaced by AI itself. The pressure to keep up has never been greater. Yet, the human brain wasn’t built for this level of ongoing mental strain.

So, how can we design learning that truly works?

What Is Cognitive Load, and Why Does it Matter?

To acquire long-lasting knowledge, the brain must successfully go through two key steps: encoding and consolidation.

1. Encoding is the process of transforming information into a format that the brain can store. This process is facilitated by working memory, which provides a mental workspace for holding and manipulating information we are currently focusing on. When we actively engage with this information by repeating it to ourselves or connecting it to things we already know, we increase the likelihood of successful encoding.
2. Consolidation is the process of stabilizing and integrating that information with existing memories stored in long-term memory. It transforms something temporary into durable memories that can be later retrieved and used. Even when encoding is successful, most information is quickly forgotten unless it undergoes consolidation.

As learning designers, we can shape experiences that support both encoding and consolidation, making it more likely that new information is truly learned. One powerful way to do this is by designing with cognitive load in mind.

Cognitive load is the mental effort it takes to process information in our working memory. Think of working memory not as a spacious warehouse, but as a small bag that can only hold a few items at a time. For example, when we try to absorb a new concept during a training session while also replying to emails, tracking upcoming meetings and mentally preparing for a presentation later in the day, our mental bag — our capacity to process and retain information — quickly becomes overloaded. When the cognitive load is too high, meaning a task or learning experience demands more than our mental bag can carry, things start to spill out.

High cognitive load is the norm today, resulting in mental fatigue, fragmented attention and learning that doesn’t stick. By designing learning experiences that manage cognitive load, we can help learners focus on what matters, retain information and apply it when it counts.

The AI Paradox: Burden and Benefit

AI presents a paradox for today’s learners. On one hand, it contributes to cognitive burden, accelerating the pace of change and requiring constant upskilling on new tools, content and processes. This adds to an already overloaded mental workspace.

On the other hand, when used intentionally, AI can help alleviate that very burden. Learning designers can leverage AI to personalize learning, reduce unnecessary complexity and surface what matters most, making it easier for the brain to encode and consolidate new information.

Rather than overwhelming learners with more AI-generated content, we can position AI as a strategic partner: one that filters out noise, directs attention to what matters and helps build learning experiences that align with how the brain learns best. Let’s explore how.

The EY Cognitive Load Learning Design Framework

1. Prioritize the essential.

Focus on what matters: Learning objectives act like a packing list for a trip. Just as you wouldn’t pack snow boots for a beach vacation, you shouldn’t include content that doesn’t directly support the learning goal. When objectives are specific and concise, such as “describe what makes feedback effective,” they help sort through content and eliminate what’s unnecessary. This reduces extraneous cognitive load by narrowing the focus to what truly matters.

AI tools can help you prioritize the most relevant content by reviewing your objectives and flagging content that may be redundant or off-topic.

Use this AI prompt: “Analyze this document and highlight only the sections that directly support the learning objectives: [specify the learning objectives]. Remove any content that doesn’t align with these objectives.”

2. Highlight relevance.

Make it personally relevant: Content that feels personally relevant to the learner is more likely to be processed deeply. When learners recognize how content connects to their own goals, experiences or challenges, it becomes more meaningful. That relevance activates their intrinsic motivation, which enhances focus and supports deeper encoding.

Use this AI prompt: “Generate a ‘what’s in it for me’ introduction for managers and individual contributors learning these skills.”

Make practice realistic: Practice helps consolidate learning, but not all practice is equal. Learners are more likely to stay engaged and remember what they’ve learned when they can apply new knowledge in realistic, varied situations that feel relevant to them. Practicing in authentic scenarios not only reinforces understanding but also strengthens memory consolidation. This kind of practice supports long-term learning and improves the transfer of knowledge to real-world situations.

Use this AI prompt: “Generate three realistic workplace scenarios where learners can practice using the feedback model from this module. Each scenario should describe the context and include a specific situation that requires giving feedback.”

3. Structure for clarity.

Organize information: Our brains crave patterns and predictability. When learning flows logically and is chunked into digestible segments, cognitive load is reduced and it’s easier for our brains to digest and integrate with existing stored knowledge. Ways to organize information include chunking and scaffolding.

• Chunking is the practice of grouping information. For example, if you’ve changed a lightbulb many times before, you don’t need to remember each step individually; your brain has grouped them into a single schema: change lightbulb.
• Scaffolding is a way to prioritize and sequence information so that foundational knowledge comes first, followed by new or progressively more complex concepts. This structure can happen within the same learning session or across multiple sessions. For example, think of scaffolding like constructing a building: you start with a solid foundation before adding subsequent floors.

Use this AI prompt: “Group this information into 3-5 modules, each covering a single main idea. Suggest clear headings and subheadings. Then, sequence the modules so that the learning starts with basic or foundational concepts, and then builds toward more complex or advanced ideas.”

Reduce friction: Processing new information already requires effort. If the presentation is cluttered, confusing or overly complex, it adds unnecessary load. Reducing friction means making content easy to scan, visually clean and intuitively structured, freeing up cognitive resources for actual learning. For example, using clear fonts, highlighting key words and limiting the number of visuals per slide can reduce extraneous cognitive load.

Use this AI prompt: “Rewrite this section for a nonexpert audience, simplifying technical terms and clarifying instructions.”

Connect to familiar language and concepts: When learners can connect new information to something they already know, it becomes easier to understand and remember. This is because the brain links new knowledge to existing networks of prior knowledge and experience, known as schemas — cognitive frameworks that help organize and interpret information. Linking learning to existing schemas reduces unnecessary mental effort and allows learners to focus on what truly matters.

Use this AI prompt: “Review the following content: [insert content or describe topic]. Suggest ways to connect this content to what learners likely already know based on their persona: [describe audience persona]. Focus on using familiar language, relevant scenarios and examples from daily experiences. Your goal is to make the content feel relatable and easy to understand by linking it to their existing knowledge and experiences.”

4. Design for attention.

Incorporate restorative breaks: Sustained attention is limited. Our brains’ prefrontal cortex (PFC) — an area especially evolved in humans that is responsible for attention, critical thinking and decision-making — easily fatigues. Without breaks, learners lose focus and learning suffers.

Restorative breaks are essential for maintaining attention, but there’s no universal formula for how long or how often they should be. The ideal break depends on many factors, such as prior knowledge, expertise, individual differences, learning modality and how much mental effort the learning requires.

Activities that evoke fascination, comfort, joy or physical movement can help restore cognitive function and reduce mental fatigue. Know your audience and adjust breaks based on their needs.

Use this AI prompt: “Review this two-hour workshop agenda and recommend where to insert breaks to maximize attention based on the audience’s needs [insert description].”

Designing for What Matters

Designing with cognitive load in mind isn’t about making learning effortless — it’s about making it purposeful. The goal should be to help ensure that learners’ limited cognitive resources are spent engaging with meaningful content, not wasted on navigating poorly designed learning experiences.

While AI can contribute to an overwhelming cognitive load, it also offers powerful tools to reduce it — if used strategically. By combining behavioral science with the strategic use of AI, we can create experiences that are not only brain friendly but also relevant, efficient and impactful.

The views reflected in this article are the views of the authors and do not necessarily reflect the views of Ernst & Young LLP or other members of the global EY organization.