2026 Learning Predictions: Where Are We? Part 2

October 1, 2026

8 min read

2026 Learning Predictions: Where Are We? Part 2 — New Africa/Shutterstock.com

October 1, 2026

8 min read

Zsolt Olah Intel Corporation

+33

Zsolt Olah Intel Corporation

+33

Overview: Do you have a seat at the table by now? Is your seat constantly moving as well?

Summarise this page with your favorite AI assistant

Learn From AI. Do Not Let AI Learn From You Only.

In the previous article I covered the first four L&D predictions for 2026 about AI, personalization, workflow learning, and skills data. I also attempted to rate the industry on the maturity index. In this article I'm exploring the next four predictions, all pointing at the same undercurrent: the future role of L&D and the change in how things get done. Of course, in 2026, AI is influencing all predictions. More on that later in the article. Here is a reminder of the rating systems we established in the previous article as the maturity level of adoption index:

Level 0. None
No known use and no observable evidence that the practice exists in the organization.
Leve 1. Exploration
Ad hoc experimentation by isolated individuals or teams, with no coordinated implementation.
Level 2. Pilot
Limited, structured trials with defined objectives, participants, and use cases, but no organization-wide implementation.
Level 3. Operationalization
The practice is routinely used in at least one business or learning workflow. Basic processes, ownership, support, and governance exist.
Level 4. Scaling
The practice is deployed across multiple functions or business units and integrated with relevant systems and processes.
Level 5. Optimization
The practice is institutionalized, produces measured value, and is continuously improved using outcome data. The organization can reproduce successful results predictably.

Let's explore the next two predictions in the list.

L&D Becomes A Strategic Capability And Performance Function

My rating: 2.

Business Outcomes Replace Activity Metrics As The Preferred Evidence

My rating: 2.

I'm pairing these two because in practice, they're often two sides of the same table (okay, it's a coin but you'll see why I chose the word "table"). The first is about the position of L&D based on the value it brings to the table. The other is the way L&D lays evidence on the table. Yes, the seat at the table.

One of HR's prominent metrics for an organization is employee engagement. Part of employee engagement is about opportunities to learn and grow. Which means L&D's strategic capabilities and influence on skills development, career movement, and performance is part of this metrics.

In 2025, global employee engagement declined for a second year to its lowest level since 2020.
- Gallup [1]

We need to wait for the 2026 results, but the trend is not looking good. There is some data in 2026 about L&D's influence, but it is not flattering either. According to research covered by HR Director, only 29% of L&D teams consistently gather and report business data and KPIs, and just 11% are consistently focused on skills growth, even though upskilling is the top priority for 56% of organizations over the coming year [2]. A separate piece on enterprise metrics puts a similar number on confidence: only 29% of L&D leaders feel confident proving ROI [3].

Measuring What Matters

Speaking of measurement. In a webinar I delivered on the topic, I illustrated the situation with a story. You might have heard this before in some shape or form:

It's late at night in the city, on a dark street. A tipsy person is on the sidewalk searching for something under the streetlight. A passerby comes and asks the person if he has lost something. The man says: "Yep. My car keys." and with that continues the search on the ground. The passerby jumps in to help. After a good 20 minutes covering every the whole area, the passerby asks: "Are you sure you lost them here?" The tipsy man looks up: "Here? No. I lost it near the football fields across the street."–and continues the search.

Puzzled by the answer, the passerby asks: "Then why are you looking for them here?"

The tipsy man points at the light: "Because the lighting is much better here."

I'll give a minute to reflect on how this joke has anything to do with L&D and measurement. Most teams are still reporting completions, hours, and satisfaction scores because those are the numbers the LMS hands you for free. We search where the light is, not where the value is. It's much harder to collect data outside of what we control.

Is it a tool or technology problem? I don't think this is a tooling problem anymore. The dashboards exist. What's missing is data literacy and the connection between a learning activity and a business result. We must start with the end in mind: when working with finance, operations, or sales we need to start with what they care about. Then, we can build backward from there, rather than starting from a course and hoping it eventually shows up in a P&L somewhere.

RedThread's 2026 vendor research backs up the direction of travel here too: analytics and measurement is now the top functionality vendors build toward, at 83% of the market, up from just 33% in 2023 [4]. The tools are catching up to the ambition. The muscle of asking "what changed" instead of "what did we deliver" is still being built inside most L&D teams.

Human Coaching, Reflection, Judgment, And Leadership Gain Importance

My rating: 3

This one I feel better about, and the data agrees it's real rather than wishful thinking. Degreed's early 2026 analysis of learning activity found that seven of the top ten skills professionals actively sought out were human or business skills, with leadership and communication in the top two spots, ahead of Excel, Python, and data analytics [5].

Udemy's business data tells a similar story from a different angle: adaptive or soft skills learning grew 25% year over year, with critical thinking up 37% and decision-making up 38% [6]. People are reaching for judgment and coaching content at the same time as they're racing to learn Copilot. At least, this was my thinking until I read this from David Rock:

Overwhelmed leaders are relying on increasingly sycophantic AI tools that reward certainty and avoid challenge and can escalate conflict [7].

The article literally suggests that there is a leadership and judgement crisis coming. Yay (sarcastic)! In theory, the more routine work AI absorbs, the more the job becomes deciding what to do with the output, which is a judgment call. But if AI reduces our ability to make judgements, it does sound like a crisis.

Illustration: while I was writing this article in Microsoft Word, Copilot jumped in (without even being asked) and informed me that it rewrote the paragraph below for clarity:

I rewrote the selected paragraph for clarity, conciseness, grammar, and flow while preserving its original meaning.

And of course, adding em dashes, sensational statements, the "it's not X, it's Y" structure, and words I do not ever use. By default.

Cognitive Offloading

The argument is that we offload routine tasks to focus on higher value activities. Remember when the autocorrect feature came out? Spelling doesn't count. Now we have time to focus on sentences and the message. Now we're offloading the creation of sentences to save time for higher value activities. At what point do we offload all of your work without learning anything?

My advice: take advantage of AI but delegate smartly. Take time to reflect on changes you agree on; learn from the interaction. Do not let AI learn from you only! And be intentional about that "time saved for higher level activities" to make sure you don't just create more work for others in the workflow who need to deal with your AI slop.

Back to trends in L&D. Digital coaching is gaining momentum because it is scalable and affordable, at least from a technology standpoint. Coaching does not rely on content delivery, so it does not compete directly with AI. I rated this a 3 rather than higher because demand is strong, but adoption requires changes to existing workflows, organizational infrastructure, policies, and probably culture. Specifically, culture that supports protected time for reflection by leadership who treats reflection as work rather than a break from it. If the leadership crisis is real, however, we may be moving in the wrong direction.

Static Catalogs, Generic Content Libraries, And LMS-Centered Experiences Lose Influence.

My rating: 4.

This is probably a 3 but should be a solid 4 by now. The RedThread number from the first article belongs here too: content creation dropped from the top vendor functionality to fifth place, and single-purpose content platforms have nearly disappeared from the market, down from 27% of vendors in 2023 to 1% today [4].

Vendors aren't building bigger libraries anymore. They're building measurement, skills infrastructure, and integration into other systems, because that's where the market has told them the value now sits. The struggle for owning skills data in the organization is real: whoever controls skills data has the advantage to build AI services on top. But for that, the data must be valid, reliable, and up to date. The old ways of HR-created job architecture-driven skills taxonomy can turn into a bottleneck due to its rigidity.

What's Changing About Content Libraries?

So, is it changing? How fast? The reality for learners might be slower. A shrinking library and a declining LMS are not the same thing. Plenty of organizations still send most learning needs to the LMS because it remains the system of record, the compliance tracker, and the place managers first look. The content inside it is getting less generic and more targeted, which is real progress, but the LMS as the front door hasn't gone anywhere yet.

Conclusion

Put all eight scores together and you get a median maturity index sitting around "operationalized in pockets, not yet the default way of working." Is that your view of your own L&D in 2026? Or are you ahead of the curve? Right now, you may feel overwhelmed because everything is changing at once: the workflows, technology, and underlying processes. You don't have to figure out everything but you need to be directionally correct to evolve.