7 predictions for L&D in 2026 with James Cranwell

2026 is just around the corner, and as 5app’s Head of Product, I’m constantly thinking about what’s coming next in the world of L&D. In truth, AI is evolving so quickly that anything could happen in the next year, but that was never going to stop me from making my L&D predictions for 2026!

It won’t surprise anyone to learn that AI will continue to dominate the changing L&D landscape, but what might be surprising is the sheer speed at which things are about to be transformed. The way things are done today is likely to feel archaic this time next year, and that’s all down to the power and efficiency of AI.

So, from artificial superintelligence to synthetic practice and AI sentience, let’s get into it!

Prediction #1: Agentic learning operations

The 2026 L&D prediction I’m most confident about, and likely the most important prediction on this list, is the sharp rise in agentic learning operations. Put simply, this means that AI agents will actively perform tasks rather than passively providing information. 

This will involve orchestrating ‘armies’ of specialised AI agents, such as planners, researchers, coaches and assessors, which will interact with a central AI orchestrator to fulfil user requests. It will drastically change the way users interact with platforms, and a key part of this prediction is that we’ll be seeing more of these ‘armies’, comprising specialist AI agents, rather than expensive, all-encompassing agents that can ‘do it all’, but not as efficiently. It’ll work just like your browser extensions – you have multiple extensions for different tasks, not just one extension wearing 15 different hats.

Agentic AI will replace manual clicking and dragging in online platforms within the next two years – yes, that includes learning management systems! Mindset, the tech partner behind VeeCoach, has its Agent Management Studio (AMS), which is a good example of a platform already implementing agentic capabilities. It allows users to simply tell the agent what they want, and the process happens behind the scenes with no need for manual setup. I can’t say too much yet, but an agentic future might just lie in wait for Helix…

A note on the stages of AI development:
Before we go any further, it’s worth covering the stages of AI development. We’re currently in stage 1, but the ones we know about right now are:

• Artificial Narrow Intelligence (ANI): AI designed to perform a single specific task or a narrow range of tasks with high efficiency, such as image recognition or language translation. It’s capable of progressing further, but is likely to hit a wall due to its neural network design.
• Artificial General Intelligence (AGI): AI with human-like cognitive abilities, reasoning and ‘emotions’, capable of understanding, learning and applying knowledge across a wide range of tasks. This could be with us as early as 2027.
• Artificial Super Intelligence (ASI): A hypothetical AI that surpasses human intelligence in every domain, including creativity, reasoning and problem solving.

Why it matters to L&D right now

AGI and ASI could be with us sooner than we think. That’s why it’s so important to use AI responsibly. The potential is huge, which is why L&D professionals need to keep their fingers on the pulse and familiarise themselves with ethical AI principles.

Prediction #2: Skills intelligence as an enterprise data product

Don’t worry – none of my other predictions are quite as terrifying as my first!

Skills intelligence (yes, like Helix) will become a vital enterprise data product – especially AI skills intelligence. Over 2026, I expect to see us moving beyond self-reporting and assessments to verifiable skills claims based on data points like LMS completions, CRM data and meeting transcripts.

This approach to skills intelligence is likely to become mainstream, enabling businesses and L&D teams to leverage real data for deeper insights into employee capabilities. The good news for Helix users is that you’re getting a head start with this exciting technology. Helix is one of the very first AI skills intelligence solutions to hit the market, and you can get started today to get access to rich employee skills data.

Prediction #3: Outcome-first L&D metrics

My next L&D prediction for 2026 is that we’re going to see a shift from vanity metrics (LMS logins, course completions, training attendees) to outcome-first L&D measurements, focusing on real business impact and ROI.

Instead of the basic surface-level metrics, we’ll be looking at new metrics such as time to competence (how long it takes someone to become competent in a skill), intervention uplift (based on learning prompted by nudges vs completely organic learning) and skills velocity (net skills verified by person) that will become standard as learning technology evolves.

Again, while I can’t say too much right now, let’s just say that tools like Helix might be able to deliver these sophisticated metrics in 2026, helping L&D teams demonstrate real, measurable ROI for their learning initiatives…

Prediction #4: AI co-facilitation for live training

In recent years, we’ve seen live AI language translation appearing in virtual instructor-led training (VILT) sessions. I predict we’ll soon be seeing AI co-facilitators appearing in these live cohort-based sessions too, where they’ll assist human facilitators and hosts by answering questions, suggesting interventions and simultaneously joining multiple breakout sessions to guide discussions.

With the advent of AGI, we could even see AI agents leading entire VILT or coaching sessions (with the right guardrails and quality assurance in place). This will significantly boost scalability and efficiency, and will be useful for localising training sessions across markets.

I expect leading live events platforms to be developing this functionality right now, with the first solutions emerging in 2026. This will help reduce headcount for live VILT sessions and make them more efficient and cost effective.

Prediction #5: Synthetic practice at scale

Elearning and live training sessions are great, but for many skills, it’s not quite enough. There’s an increasing appetite for high-fidelity scenario simulators and training sandboxes for safe experimentation, which is also known as synthetic practice.

This goes further than simple roleplay exercises. In 2026, I predict we’ll start seeing multi-agent simulations that can replicate complex team dynamics or even a company’s entire financial structure and customer base. This gives learners a safe place to try new things and learn from their mistakes, providing immersive and personalised learning experiences.

In its most advanced form, we might see these multi-agent simulations integrating with augmented reality (AR) or virtual reality (VR) technologies for the most realistic synthetic practice. Imagine the potential of these solutions for things like helping an entire team of employees escape a burning building as part of a fire drill, or safely locking down a bank in light of an attempted robbery. It’s a safe way to learn the ropes, and it’ll certainly make the whole experience more memorable!

Prediction #6: Privacy and ethics in AI governance

I know, I know – AI governance may not be the most exciting L&D prediction on this list… but it’s certainly one of the most important.

A key barrier to adopting AI solutions is the fear that user and organisational data won’t be secure, and as AI becomes more prevalent in the L&D industry, ‘privacy by design’ and baked-in data governance will become non-negotiables. 

AI is obviously going nowhere, so as a response to the nervousness around data privacy and AI ethics, we’re going to see a huge uptick in ‘red teaming’ – a way to detect vulnerabilities like data leakage or biases in LLMs. Red teaming works a lot like software penetration testing – by running adversarial prompts, designed to ‘trick’ AI tools into giving unwanted responses, teams can test the robustness of their AI software. This will force AI vendors to up their game, and give organisations the peace of mind they need that their next-generation tools and platforms are safe and fit for purpose.

Prediction #7: Composable and interoperable learning stacks

72% of learning teams are running three or more learning systems in 2025, and that number is only going to grow as more AI-powered platforms and products hit the market. That’s why my final prediction is around the rise of interoperable learning stacks, where platforms will be expected to seamlessly integrate and exchange data. Think of it like the Lego of learning – any Lego set you buy will connect with any other set, giving you complete control and freedom over the way you build your learning technology ecosystem.

Universal compatibility can be achieved either through all software becoming fully open source (which will never happen), or realistically, through open APIs. Learning platforms offering easy integration through open APIs will thrive and survive, even in an increasingly competitive market, as L&D teams won’t need to switch up their entire stack every time they want to swap out one tool.

Also – you guessed it – I’ll be able to share more on this in the coming months, but we’re currently working on a general API to allow Helix users to integrate our AI skills intelligence solution with their other platforms, allowing for a more flexible and customised learning ecosystem, all powered and informed by real-time skills data.

Let’s see what 2026 has in store…

I’m excited to see which of my predictions come true in 2026 – but what I can confirm is that our product team is working hard to encompass as many of these points as possible in the next iteration of Helix, as well as turbocharging our LMS, The Hub, and working with our tech partners to ensure VeeCoach and VeeCreate keep up with the 2026 L&D trends.

Until then, why not get ahead of the curve and give Helix, our AI skills intelligence solution, a try? We’re proud to be one of the very first to market with this innovative solution, and there’s nothing more satisfying than being one of the first to get to grips with a new technology!