Our Take
Cornerstone is executing a real platform pivot, but it remains an overlay on legacy LMS infrastructure while pure-AI competitors like Sana already ship content-generation-first from the ground up.
Why it matters
L&D budgets are shifting away from traditional course-building platforms toward AI-native enablement at speed. Cornerstone controls 7,000 enterprise accounts and $1 billion in revenue; how it plays this transition will signal whether incumbents can outrun startups in the skills graph space.
Do this week
L&D leaders: audit whether your current platform's renewal roadmap includes AI-native content generation and open people-graph APIs before Q3 renewal cycles lock in.
Cornerstone launches Workforce AI, a skills and agent layer atop its learning platform
Cornerstone, the $1 billion revenue leader in corporate learning technology (company-reported), announced Workforce AI this week. The platform integrates three components: skills and capability inference, an open "People Graph" that ingests data from HRMS, LMS, project systems, and employee surveys, and packaged "Readiness Agents" that deliver role-specific development or staffing recommendations without manual intervention.
The company has deployed this to at least three named customers with measurable focus. A wealth management firm is using the platform to identify and develop high-performing wealth managers through signals previously unavailable in traditional talent management. A media company navigating a merger is using it to map tacit skills pools and identify at-risk talent for retention programs. Cisco is applying skills intelligence to project-based staffing across its telco operations. A healthcare company reports projected savings of $30 million to $40 million (company-reported) from an "Embark Navigator Agent" that personalizes development plans for critical roles.
The platform represents a strategic shift for Cornerstone, which has spent 27 years as the dominant incumbent in SCORM-based course authoring and learning management. Since going private in 2021 for $5.5 billion (company-reported), it acquired EdCast (content recommendation), Skyhive (labor market data), and Tailspin (VR). Workforce AI consolidates these into a single architecture. Content generation and LMS administration agents are already live; open APIs for custom agents are coming.
The published-course market is collapsing; AI-native enablement is where L&D budget is moving
Until late 2022, corporate L&D competed on the depth and breadth of pre-authored courses and content libraries. A single course could cost $50,000 or more to develop, taking 6 to 9 months. That process is now economically obsolete. AI-generated courses can be authored in days, edited at low cost, and delivered as conversational enablement rather than linear chapters. Users ask questions and receive contextual answers instead of paging through modules.
The shift is not hypothetical. Josh Bersin, who is the source of this reporting, notes that the traditional published-training market (worth roughly $400 billion) is being reinvented at speed. Newer vendors like Sana (backed by Workday), Docebo, Absorb, and others are already shipping AI-native platforms. OpenAI has begun integrating content from Coursera. Anthropic is embedding its engineers into customer deployments to build similar systems from scratch.
The consequence for incumbents is clear: L&D teams are losing budgets and seniority as they stop building 9-month courses and start acting as enablement partners embedded in the business. Cornerstone's installed base of 7,000 enterprise customers and 140 million users (company-reported) makes it a powerful incumbent, but only if it can migrate existing revenue to AI-native tiers before churn accelerates.
Validate whether Workforce AI closes the gap with pure-AI competitors
Cornerstone's Workforce AI is not yet a pure AI-native platform in the way Sana (built from scratch on AI inference and content generation) operates. It runs as an overlay on Cornerstone Galaxy, the legacy learning platform, and packaged agents are not yet open for custom tuning.
For practitioners evaluating renewal or net-new solutions, the question is whether the open People Graph and agent framework mature fast enough to match the velocity of startups that have no legacy LMS to maintain. Request pilot access to the custom-agent roadmap and labor market data integrations (Skyhive) before signing multi-year contracts. Confirm whether content generation features ship at the pace your business requires to compete with teams using pure-AI tools like NotebookLM or platforms designed for rapid iteration rather than compliance and governance workflows.