TalentLMS Learning Debt Report: AI Use May Be Masking Skill Gaps

TalentLMS’ Learning Debt Report finds 59% of employees use AI tools to perform tasks they weren’t trained for, with 37% saying AI makes them appear more competent than they are and 29% delivering work they can’t fully explain. Employees who fall behind on skill development are nearly 6x more likely to make preventable mistakes, and 65% report output quality suffers. Overall, the survey suggests AI boosts short-term productivity but increases the risk of hidden skill gaps unless training keeps pace.

Analysis

The market should treat this as a slow-burn operating issue, not a near-term macro shock. In the next 1-2 quarters, AI is more likely to inflate reported productivity than to reduce error rates, which means the first visible effect is probably margin leakage from rework, QA, customer escalations, and manager oversight rather than a sudden revenue hit. That makes AI-augmented workflows a mixed blessing for services-heavy businesses: they can hold headcount flat, but the hidden cost of doing more with less is lower execution quality.

The clearest beneficiaries are enterprise learning platforms and workflow-training software that can sit inside the job process, not static course libraries. Public comps with the best setup are Docebo (DCBO), Workday (WDAY) learning modules, and SAP/Oracle HCM adjacencies; weaker are commoditized content sellers like Udemy (UDMY) and Skillsoft (SKIL), where AI can substitute for generic knowledge faster than it can substitute for role-specific training. The second-order risk is that buyers initially delay spend because AI masks the gap, then come back only after mistakes appear in compliance, customer support, or internal controls.

Contrarian view: the consensus may be overestimating how much of this converts into incremental L&D budgets. If management teams believe copilots can bridge skill gaps cheaply, training budgets could stay flat for longer than bulls expect, and the real capital budget may go to AI tools, not training systems. The thesis is falsified if enterprise AI adoption keeps rising while training/learning attach rates, L&D headcount, or rework metrics do not deteriorate over the next 2-3 reporting cycles.

Key Decisions for Investors

• Long DCBO vs short UDMY into the next 1-2 earnings cycles: DCBO is better positioned to monetize role-specific, workflow-adjacent training; UDMY is more exposed to commoditized content substitution. Risk/reward is attractive if enterprises start refreshing training budgets after Q3 guidance season.
• Accumulate WDAY on weakness as a 6-18 month AI governance/learning workflow beneficiary; the bull case is not AI hype, but that customers will need training embedded in HR workflows once error rates show up. Falsifier: no improvement in learning-module attach rates or HCM cross-sell commentary by the next two quarters.
• Keep SKIL on a watchlist for short-side setups on any post-earnings strength: if management leans on AI-driven efficiency without evidence of durable enterprise demand, multiple compression is likely. Use a tight risk box around the next guidance update.
• Pair trade idea: long enterprise software with embedded training/workflow capabilities (WDAY, SAP) against a basket of generic e-learning names (UDMY, SKIL). This expresses the view that buyers will pay for process-integrated learning, not standalone content, over the next 3-6 months.