What Happened at the DevLearn Conference 2025 and What L&D Teams Should Take From It

The DevLearn conference 2025 ran from November 12 to 14 at the MGM Grand in Las Vegas, drawing over 4,000 learning professionals across more than 200 sessions, hands-on labs, and expo showcases. If you were there, you already know it was one of the more substantive years in recent memory. If you weren’t, this recap covers the sessions, speakers, and industry signals that are most worth your attention heading into the rest of the year.

The opening keynote set a clear tone: L&D teams need to prove their value or risk being sidelined

Dr. Keith Keating’s opening address was the right way to start a conference that has increasingly had to reckon with its own identity. As Chief Learning and Development Officer at BDO Canada and author of “Hidden Value: How to Reveal the Impact of Organizational Learning,” Keating brought a bold roadmap for transforming learning into a strategic powerhouse that seeks, creates, and amplifies organizational value. What made the keynote resonate, based on the conversations we tracked afterward, was not the polished argument but the personal framing underneath it.

Keating prefers the term “L&D practitioner” over “L&D professional,” and the distinction was deliberate. His argument: the field has a low barrier to entry, and when people perform poorly, it damages how the broader organization views learning and development long-term. That’s a harder message to sit with than the usual “seat at the table” framing, and it’s one that rings true to anyone who has watched a badly designed compliance course get deployed enterprise-wide because no one pushed back.

In our read of the room, the practitioners who engaged most with this session were the ones already doing the unglamorous work of connecting training outcomes to business metrics. The ones nodding hardest during “we are problem solvers, and L&D is one way we solve problems, not the only way” were typically people running training operations at scale, not just designing individual courses. That tension between strategic contribution and tactical delivery came up repeatedly across devlearn sessions 2025.

For context, the conference was designed for professionals and leaders passionate about leveraging technology to create innovative and impactful learning experiences, offering strategies, practical tools, and educational opportunities for decision-makers to explore AI and advance learning in their organizations.

AI at DevLearn 2025 was different this year, and that difference actually matters

The AI conversation at the devlearn conference 2025 has matured. Last year’s common question was “does it have AI?” This year the focus shifted toward problem-specific queries: does AI genuinely improve this outcome? That is a meaningful shift, and it played out across both the session rooms and the expo floor.

Yulia Barnakova from Accenture, who delivered a Thursday morning keynote on how L&D can lead in the age of AI, made the consistent message clear: AI is not here to replace learning professionals, but to amplify what they do. Her session drew from Accenture’s work across large-scale enterprise deployments and covered practical ground: how to use AI for first-draft course generation, how to build branching scenarios faster, and how to create microlearning at volume without sacrificing quality.

Barnakova emphasized that as AI changes the nature of work, L&D professionals need hands-on experience with AI tools to train others to use them effectively, efficiently, and ethically. Understanding where AI adds value and where its limitations lie is now increasingly essential for learning team members.

The pre-conference AI summits reinforced this. The AI for Learning Summit and AI 101 for L&D Professionals workshop went beyond the hype to focus on practical implementation, with sessions covering first draft generation, scenario building, and AI-enhanced microlearning that accelerate eLearning development significantly.

Where the conversations got more interesting, in our view, was the Guild Masters panel. The panel raised a question that doesn’t have a clean answer: most people using AI now already know how to do their jobs and can therefore judge if AI outputs are acceptable. But what about people who don’t already have the skills? How do they judge the output? How do we train people to judge AI output if they don’t already know how to do the job? That question is not rhetorical. It points to a genuine skill gap problem that no LMS or LXP has cleanly solved yet.

The takeaway for learning leaders is practical: start with the work, not the technology. Identify where speed, consistency, or cognitive load create bottlenecks. Build small but durable AI workflows around those tasks. And create communities inside the organization where teams can share prompt patterns, mistakes, and experiments without fear.

Training operations and scheduling emerged as a real pressure point, not just a background topic

If you spent time on the expo floor during the devlearn conference 2025, you noticed something that the stage sessions didn’t fully surface: a lot of L&D teams are struggling with the operational side of training delivery, not just content design. The conversations happening in vendor booths around instructor-led training (ILT) scheduling, resource allocation, and cost tracking were some of the most candid we encountered across the three days.

Training Orchestra, for example, was present to show how organizations can reduce costs of ILT and vILT by automating and optimizing training operations, addressing the complexities in session scheduling, resource management, instructor collaboration, cost tracking, and reporting through a Training Resource Management System. That kind of tooling doesn’t get keynote-level attention, but it addresses the friction point that actually slows down enterprise training programs. Other platforms in the training scheduling and management space, including Administrate, accessplanit, Arlo, and SimpliTrain, represent a broader ecosystem addressing these operational challenges.

The conversation around platform migration also surfaced repeatedly. Organizations are often locked into their current learning ecosystems, with large amounts of data and content representing significant investments. At conferences like DevLearn, teams discover authoring platforms and learning content management systems that would be significant steps forward, but they face the daunting challenge of migrating all that data. Too often, teams stay the course because the price of change seems too high.

This is a training management problem as much as a technology problem. Teams that have mapped their ILT calendars, instructor pools, and compliance requirements into one system face real disruption when they consider switching. The vendors with the most compelling booth conversations were the ones who led with migration support rather than feature lists.

The table below summarizes the key operational themes that surfaced at DevLearn 2025 alongside the training functions they affect:

Immersive tech, microlearning, and accessibility moved from conference trends to floor-level practice

Few things at DevLearn sparked as much excitement as the VR and AR demos. Watching participants light up during hands-on sessions was a highlight, with one attendee commenting “VR demos blew my mind today.” The future of immersive training feels closer and more accessible than ever, promising richer, more engaging learning experiences.

What made this year different from previous iterations of the “XR is coming” conversation is that the demonstrations were grounded in actual use cases rather than proofs of concept. A standout example from the Articuland sessions was a project from Ninja Tropics COO Carlos Alfaro on building a 360-degree virtual city for learning, which drew strong interest from attendees looking for ways to replicate that kind of experience without a full production studio.

Microlearning followed a similar trajectory. Microlearning emerged as one of the most actionable topics, backed by case studies that resonated deeply. Many attendees left sessions ready to implement these strategies right away, with several noting they couldn’t wait to “bring microlearning back to my team.”

Accessibility got arguably the most substantive treatment we’ve seen at a DevLearn in some time. Sessions addressed how to identify and fix the most common accessibility mistakes that cause courses to fail accessibility standards, explored designing inclusive eLearning experiences, and covered how to design for neurodiversity using Universal Design for Learning (UDL) principles, with practical strategies for supporting learners with ADHD, autism, dyslexia, and other learning differences.

The framing that stuck with us was the emphasis on accessibility as a starting condition rather than a post-production fix. If you’re building courses and running an accessibility check at the end, you are already doing it wrong, and several sessions at DevLearn 2025 made that case persuasively.

The devlearn sessions 2025 that gave practitioners the most to act on immediately

Not every session at a conference this size is worth your time, and DevLearn 2025 was no exception. But the sessions that generated the most follow-up conversation shared a common quality: they brought specific outcomes, not just frameworks.

Clark Quinn’s session on spaced learning offered an important reminder that the science of how people learn has not changed. A program succeeds when it aligns with how the brain takes in information, forms memories, and turns insight into behavior. He also cautioned against misusing terms like microlearning and “learning in the flow,” which often signal convenience rather than cognitive alignment.

Sessions exploring the 5 Moments of Need framework, developed by Bob Mosher and Conrad Gottfredson, provided structured thinking around the five distinct moments when workers require learning support: learning something for the first time, expanding existing knowledge, applying skills in workflow, troubleshooting problems, and adapting to change. For training teams that are trying to move away from event-based learning and toward continuous performance support, this framework gave practical scaffolding to do that work.

The DemoFest showcase also generated unusually strong conversation this year. One team won Best in Blended Learning Solutions for an electric utility client’s three-year technical training program, supported by a sci-fi webcomic narrative with badges and points. That kind of entry tends to open up discussion about what “blended” actually means in a high-stakes ILT context, and those conversations continued long after DemoFest closed.

Session formats themselves got more varied. The 90-minute Mastery Sessions and BYOD labs were consistently packed, which tells you something about what practitioners actually want when they spend time at a conference: they want to do something, not just hear about it.

What the community told the Learning Guild they want next, and what that signals for 2026

The Learning Guild published feedback from DevLearn 2025 influencers on what they want to see at the next event, and the signal coming back was clear in two directions: more honest failure stories, and less AI surface-level coverage.

AI continues to be a major point of curiosity, but the most interesting conversations are no longer about what is possible. They are about what is actually working. The most compelling conversations focus less on what AI could do and more on what it is already doing in practice.

Several voices emphasized the need for more stories of failure to be shared. Not everything works, and that is where some of the most valuable learning happens. These types of conversations open the door to collaboration, problem-solving, and shared understanding.

The organizations that will succeed in 2026 are not the ones that adopt the most tools. They are the ones that take a disciplined approach to performance, empower their teams with clarity and confidence, and use AI in ways that support real capability building.

That’s the real takeaway from the devlearn conference 2025, beyond any individual session or keynote. The field is moving from experimentation to implementation, and the practitioners who thrive in that environment are the ones who combine learning science fundamentals with honest evaluation of what technology actually changes in their specific context.

If your team is evaluating LMS platforms, training management systems, or eLearning authoring tools right now, the conversations that happened at DevLearn 2025 are worth more than any vendor demo in isolation. The community is the signal.