Here’s how fast AI is moving: across the five leading frontier AI labs, the median gap between new model releases has compressed from roughly 37.5 days in 2023 to just 11 days so far in 2026. Zoom out to a single month and it’s even starker: August 2026 alone saw 24 confirmed AI models launched by 18 different providers in just 31 days, reportedly the fastest release month in AI history. New capabilities are landing before most companies have finished writing the training deck for the last one.

Now compare that to how a typical L&D team ships a course: an intake call, subject matter expert (SME) interviews, a storyboard review, a recording session, a QA pass. Weeks of work, easily, sometimes months. By the time that course goes live, the tool it was built around may already have a new interface, a new agent mode, or a workflow that’s quietly gone stale.

This is the gap every learning team is living with right now, and it shows up in how unprepared people actually feel. Docebo’s 2026 AI Readiness Gap report found that 85% of learners say the training they get doesn’t help them fully understand or use AI in their role, and one in five received no AI training at all. AI skills development for employees is the single highest priority skill for both learners and leaders heading into the next 12 to 18 months. People know exactly what they need. Nobody’s built it fast enough to hand over.

The good news: this isn’t a problem you have to out-build. It’s not about hoarding content. It’s a problem you can out-source, deliberately and at scale.

Tackling the training problem(s): The case for buying off-the-shelf content

Nobody’s short on ambition. Every Head of L&D on your Slack channels is fielding requests from sales, HR, finance, and legal, all wanting “AI training” yesterday. The trouble is what happens when a small internal team tries to be the sole author of that training.

Here are the issues they face: pacing, depth, and fragmentation.

The pace problem

By the time a course on a specific AI tool ships, that tool has already added three features the course doesn’t mention. To top it all, the AI Readiness Gap report found 64% of learning leaders struggle to find the time to deliver learning at all let alone build it.

When Copilot or Gemini ships a new capability, a good third-party AI learning content catalog can have content live in weeks. Instead of building a course from scratch, you can spend that time curating one that’s already good and relevant.

The depth problem

Most teams can build one good course, maybe two, before instructional design capacity runs dry, which means AI training ends up shallow and generic instead of role-specific.

The fragmentation problem

When the central team can’t move fast enough, individual departments start buying their own point solutions, and six months later nobody can say what “AI trained” even means across the company.

AI literacy for finance looks nothing like AI literacy for developers, which looks nothing like AI literacy for a regulated healthcare workflow. No single internal SME can credibly cover that range. A catalog with multiple specialized publishers solves both the depth and fragmentation problems.

But off-the-shelf AI training content has additional benefits:

  1. Vetted content de-risks the tricky topics. For topics like AI ethics, data privacy, and regulated-industry use of AI, there’s value in content that’s been vetted by specialists, has CEUs/certifications, or is kept current by a dedicated provider.
  2. The economics simply work better. For some organizations, it can be more efficient to buy 80% of AI content (generic skills, vendor tools, certifications) and reserve internal capacity for the 20% that’s uniquely theirs (proprietary workflows, IP, customer data).

Every catalog on the market has an AI section now. Telling the good from the padded is the harder part. 

How to evaluate AI training providers: The six-point framework

Every catalog claims to have “AI content” now. Here’s what separates the useful from the padded:

  1. Substance over buzzwords. Look for scenarios, labs, and projects, not talking-head videos about “the future of work.” Level-targeting should be explicit: 101 versus advanced, business versus technical.
  2. Coverage of the full AI stack. Off-the-shelf content should not just look at literacy, prompting, or data fundamentals. That’s just the start. It should also address the use of specific tools (Copilot, Office, CRM, contact centre, etc.) and it should focus on sector- and role-relevant use cases. Sectors like Healthcare, Finance, Manufacturing, and Retail will all have different ways to leverage AI. And within those sectors, each role will also use it differently.
  3. Assessment and proof points. Off-the-shelf content providers should have quizzes and projects that show a skill changed, plus credentials or CEUs where they matter.
  4. Governance and risk. There should be content on acceptable use, data handling, bias, and regulatory considerations. Not to mention, the provider should have a track record of updating their content as laws and tools change.
  5. Localization. You might have a small team, but your company’s workforce is global, and you still need to deliver learning to them. An off-the-shelf content provider that only has languages spoken in North America will not cut it.
  6. Operational fit. Last but not least, the provider’s off-the-shelf content solution should cleanly integrate with your LMS or LXP, and it should include searchability and metadata, allowing admins to curate the content fast.

Examples of off-the-shelf content partners and how they help

Different off-the-shelf content providers will offer different courses. Let’s take a look at OpenSesame and Go1 to see what you can expect.

Go1 

Go1 gives people baseline AI fluency, practical ways to experiment inside their own workflows and clear guidance on responsible use. That coverage comes from three places: 

  • Breadth across roles and industries: AI literacy for the whole company, plus role-based content for HR, sales, managers, etc.
  • Fast discovery across many providers in a single catalog (one search surface for “AI for managers,” “Copilot in Excel,” “prompting for marketers,” etc.).
  • Blended programs that mix AI with adjacent skills (change management, digital productivity, communication).

Go1’s CEO, Chris Eigeland, has seen the impact that their off-the-shelf content has in training, noting “The companies that are most progressed through their AI journey are also those that promote significant ground-level experimentation and curiosity.” 

5 course recommendations: AI in the Workplace (Ethena), AI Privacy Essentials (Kineo), AI Ethical Use and Responsibility (Intellezy), Avoiding AI Workslop (Simon Sez IT), AI Prompt Writing–Beginner (Intellezy)

OpenSesame 

OpenSesame takes people past AI basics into applied skills, credentialed programs, and the judgment calls that come with multi-step AI work. The depth comes from three places: 

  • Applied AI and data skills with structured learning paths and often lab- or project-based, formats that allow you to track whether skills have been acquired
  • Certification and CEU-bearing content (e.g., for technical, compliance, or professional-body-aligned programs).
  • Publisher diversity: the ability to pick from multiple specialized publishers (e.g., data/analytics houses, security providers) without sourcing each one individually.

As Kyle Gagliardi, Technology Curation Specialist at OpenSesame, puts it: “Because AI is changing so quickly, no single course or publisher can cover the full learning need. Combining different perspectives allows organizations to address tool skills, workflow design, strategic thinking, and the human side of adoption while keeping the program flexible as the technology changes.”

5 course recommendations: AI Is a Thinking Partner, not a Decision Maker (Mindscaling), Introduction to Claude (Simon Sez IT), Design Your Agentic AI Workflow (GoodHabitz), Claude – Claude Cowork for Automating Tasks with Connectors (Mind Channel), Preserving Human Judgment in an Automated World (MIT Sloan Management Review) 

None of this replaces what you build in-house. It clears the ground so your team’s time is spent on content only you can make. Tools like Docebo Creator let you turn company-specific workflows and policies into lessons fast, once the generic groundwork is already covered by a partner catalog. The winning programs aren’t off-the-shelf or homegrown. They’re both, stitched together on purpose.

How to get started in 90 days

In one quarter, you can have one audience trained on a blended path, and the before-and-after numbers to provide it worked. Here’s the sequence. 

Day 1 to 30: pick your audience and set a baseline
Start with the team feeling the most AI pressure right now, not the whole company. That could be sales, customer support or a finance team suddenly asked to automate reporting. Then audit what AI content you already own; there’s a good chance you have the answer, and the gaps are smaller and more specific than “we need more AI training.” Write down what they can do today: tool adoption, self-reported confidence, one or two tasks you expect to get faster. The baseline is the part teams skip, and it’s the only reason day 90 produces evidence instead of anecdotes. 

Day 31 to 60: shortlist and build the path
Run the six-point framework against partner content in Content Marketplace to shortlist what fits. Connect Docebo’s MCP Server to your AI assistant and you can compare and curate that content in a conversation instead of clicking through catalogs. Then blend it: off-the-shelf content covers the foundations, and Creator turns your own workflows, policies, and examples into lessons your learners won’t skip.

Day 61 to 90: launch and measure
Roll it out to that one audience and check it against your baseline. Two metrics is plenty. What you get at the end isn’t just a trained team; it’s a repeatable pattern and the internal proof to run the next audience in weeks. 

Closing the gap 

The pace of AI isn’t going to slow down to match your build cycle. But the gap between what your people need and what you can hand them is a sourcing problem, not a capacity problem, and that one you can solve this quarter. 

Learn how Docebo’s Content Marketplace puts off-the-shelf content to work alongside your own.