AI is already reshaping L&D from the bottom up
AI in L&D has moved from experimentation to expectation. 87% of teams are already using it, and only 2% have no implementation plans. If you wait, you will be late.
This article provides evidence of this tipping point, what it means for your team, and the moves you should make now. So you can scale with confidence.
About research
The insights in this article are based on the AI in Learning & Development Report 2026, a global study conducted by Synthesia in collaboration with Dr. Philippa Hardman. The survey received 421 responses from L&D leaders, instructional designers, learning technologists, HR and talent teams, and subject matter experts in North America, Europe, APAC, Latin America, and MEA.
Participants represent a wide range of industries, including technology, education, consulting, manufacturing, healthcare, finance, and government, with a strong skew toward companies, with nearly half working for organizations with 1,000 or more employees.
The numbers don’t lie: AI adoption in L&D is now mainstream, but not from the usual sources
With 87% of L&D teams actively using AI, we’ve moved away from the “should we?” mindset. According to the “How Fast Can You Scale?” study, 57% of teams are actively using AI in production, and the remaining 30% are running pilots. Compare this to last year, when 20% of organizations were not using AI at all. Currently, only 2% have no implementation plan.
“Is your L&D team currently using AI tools in its learning and development programs?”
I have seen this change happen firsthand. A year ago, the team was quietly testing ChatGPT for course outlines. Most AI adoption was bottom-up and not aligned with formal organizational strategy. Now those same teams share AI playbooks and team workflows.
The driving factors are clear. Cost and time savings, faster production, and efficiency.
A learning team within a global pharmaceutical manufacturer started with zero AI capabilities and relied entirely on manual processes for goals, scenarios, and assessments. Within six months, we built a documented library of prompts and workflows that standardize how AI supports every step of the design process. The breakthrough happened when we stopped treating AI as a content shortcut and rebuilt our workflows around clarity, templates, and quality control. This shift cut development cycles in half and reduced rework because the team finally had a shared way to produce consistent output. This is very important to do correctly. Transformation came not from experimenting with tools, but from institutionalizing repeatable AI practices that scale across the organization.
While executives discuss governance, teams are already using AI to generate quiz questions (60%) and text-to-speech narration (63%) at scale. This bottom-up implementation changes the way you approach your AI strategy. It’s also showing up in administrative work, where teams rely on AI to draft reports, policies, and internal communications that used to take entire afternoons.
What to do now:
Share these adoption statistics with your executive sponsor to secure budget. Create a formal team AI playbook that documents your current use cases and workflows. Set a 90-day scaling goal to scale your pilot to defined team-level workflows.
From faster content to smarter learning: How the use of AI has evolved
Early AI learning use cases focused on speed of content generation, such as creating quizzes, writing scripts, and translating content. Saving time remains important to 88% of teams. But that’s no longer the case.
The team is moving to solutions aimed at improving learner experience, not just speed, such as adaptive pathways, skill mapping, and AI tutors.
A large consumer services company struggled with disparity in competency across its customer-facing team, but lacked the bandwidth to create targeted training at the pace the business needed. This shift occurred as we moved from relying on text-based training documents to using AI to generate training videos, guides, scenarios, and refreshers directly tied to skill gaps that supervisors perceive in real-time. Managers can refine these AI videos in minutes, enabling continuous upskilling without taking employees off the floor. Improved consistency of performance across locations without increasing headcount. This is very important to do correctly. The real leverage lies in accelerating content creation to address emerging gaps, rather than through advanced automation.
The next frontier is for teams to analyze needs with AI, design adaptive sequences, implement tutors, and evaluate with predictive analytics. 88% now value time savings, while 72% expect personalized learning to be a key benefit.
What to do now:
Choose two or three use cases other than content creation: assessment or simulation, adaptive pathways, or AI instructors. Create or surface assessment rubrics designed around variable outcomes, such as reducing error rates or time to competency. Pre-define success metrics so you can prove value beyond time saved.
The new reality for L&D professionals: Strategic architects, not content creators
The purpose of modern L&D is changing. Its role is no longer to drive the course, but to design workforce capabilities and align skills, systems and communications with business priorities.
In reality, most teams still spend most of their time on production tasks, such as revising slides, drafting scripts, rewriting SME content, managing revisions, and dealing with endless admins that can take weeks.
This is an important gap. Although the intentions are strategic, the day-to-day reality is executive, reactive, and dominated by low-leverage work. AI now handles mundane tasks such as drafting, translating, and basic evaluations. This allows L&D to focus on capability building and learning architecture.
But it requires new skills. For example, 67% of L&D professionals want AI skills training for their teams, and 63% need guidance on integrating AI into their workflows.
“What kind of training or support can help your team use AI more effectively in L&D?”
Effective L&D teams are more like performance consultants than course builders. Connect business strategy to employee capabilities, discover skills gaps early, and measure impact on outcomes rather than completion.
On the ground, AI also takes over the administrative burden. Draft stakeholder updates, SOPs, and policy updates to keep your team focused on design and change management.
Human judgment remains essential. AI can generate content, but humans make sure it fits the culture, context, and behavior change goals.
What to do now:
Map team roles to new skill needs such as AI literacy, data fluency, ethical implementation, and systems thinking. Provide each team member with at least 5 hours of role-specific AI training. Establish an internal AI community of practice to share prompts, workflows, and quality standards.
Training Synergy: The Secret Weapon of AI Deployment
AI adoption doesn’t increase just because teams have access to tools. This rate is increasing as teams receive intensive, small-scale, targeted training that is directly tied to their work. This pattern is consistent. Usage increases as teams learn the basics, spread best practices (and workflows), and immediately apply them to real projects.
In my experience, effective training is lightweight, applied, and role-specific. Skip general prompts and share how specific prompts and workflows can improve the work already being done within the systems your team already uses.
What to do now:
Provide short, practical suggestions that focus on real-world jobs and tasks. Create a small set of shared prompts and templates tied to your core workflow. Track simple before and after metrics. “With the new workflows, how much time did this task take you last month versus this month?” Review these workflows on an ongoing basis.
Measuring what matters: Moving from speed to results (or new behaviors)
Teams are moving from time savings (88% currently) to business impact (55% expect) and personalization (72% expect), but 63% need help measuring impact.
Most teams can quantify the time saved. Few people are able to connect AI to results. But if you start with known information, you can sue.
The communications training team I worked with reduced onboarding time, but they couldn’t directly link that reduction to business outcomes. Using a simple ROI model, we traced the ramp time savings back to hard numbers.
New employee salary Training content development costs Supervising costs per new employee Expected ramp-up time for new employees
After implementing new AI-enabled content and workflows, onboarding was reduced from 26 weeks to 7 weeks.
This is very important to do correctly. Business leaders don’t care about “fast” onboarding. They value faster onboarding, which reduces the cost of training new employees. This frees up supervisor capacity and speeds up productivity. The real value lies in linking these results to specific AI workflows.
Successful teams establish baselines, choose metrics that business leaders care about, and create consistent reporting.
What to do now:
Select one program to measure end-to-end. Choose KPIs that can connect AI inputs to business outcomes. Start building your case by running a 90-day experiment and sharing your results with stakeholders.
The future of Agentic is already here (and our L&D team is building it)
49% are considering AI tutoring, 43% are considering AI-powered coaching, and teams are currently building autonomous learning systems.
“Which of the following agent functions are you considering for your L&D operations?”
What is agent AI?
Agenttic AI is an L&D term that refers to AI systems that perform goal-driven actions. Guide learners, make decisions, and adapt interventions without continuous human prompting.
Consider an AI tutor that detects learner struggles, adjusts difficulty, recommends resources, and schedules follow-up assessments.
I’m seeing this in the field. An e-commerce furniture retailer uses an AI coach to analyze agent performance and coach them on specific behaviors. A healthcare company built an AI FAQ agent that answers technical questions, suggests solutions, and routes complex questions to human experts.
Where should these agents live? 27% of L&D professionals don’t know. Opinions are divided between LMS, productivity tools, standalone apps, and integration layers. Only 47% believe the LMS will remain the backbone.
It’s also difficult to integrate. 50% of teams require technical support to connect AI tools to existing systems, especially if questions are raised in the IT council.
What to do now:
Try one agent use case in a low-risk area like onboarding or product knowledge. As a first or second step, design guardrails for agent instructions. Raise IT concerns early and often before proceeding with the pilot.
final thoughts
The turning point has not yet come. Here it is.
AI is already reshaping L&D from the bottom up, and the teams that move fastest are not the ones with the biggest budgets or the most mature strategies. They are people who focus on three things. One is building small but meaningful skills, rewriting workflows that AI can actually leverage, and measuring outcomes that leaders actually care about.
This is very important to do correctly. The value of AI in L&D is determined by the number of tools you deploy. Rather, it depends on how effectively you can reduce operational friction, accelerate capacity building, and demonstrate effectiveness in terms of performance, quality, or speed of productivity improvement. The transition from content creation to feature design, activity metrics to business outcomes, and experimentation to scalable practice is already underway.
The teams currently winning are:
Invest in lightweight, role-relevant AI skills rather than large training programs. Build clear, repeatable workflows where AI handles low-leverage work. Anchor any use case to key business metrics. Pilot small experiments with equipment and scale only those that have proven value.
Start with one high-impact workflow. Prove your results. Expand intentionally.
In this way, L&D moves from experimenting with AI to reimagining how organizations learn, perform, and adapt.
Synthesia
Synthesia is an enterprise AI video platform for L&D and communications teams. Create, translate, and update training videos in minutes with studio-quality avatars, accurate lip-sync, and governance controls built for global organizations.
