News

Ryan Austin is founder and CEO of Cognota, a purpose-built operating system for LearnOps, Talent Ops, and HR Ops.

getty​Every enterprise software category eventually gets an operating system—one platform that becomes the default place work happens, with everything else plugging into it. Salesforce did it for sales. Workday did it for HR records. The open question is who will do it for workforce enablement, the group of functions responsible for how people grow, perform and get work done—and what role AI will play in the tech stack.

Today, workforce enablement is three separate teams running three separate stacks. Learning and development (L&D) owns learning delivery. Talent owns performance and mobility. HR operations owns the systems of record. The fragmentation shows up in the numbers: Only about 18% of organizations with more than 1,000 employees use a shared-services model to unify HR delivery, and just 19% of core HR processes are currently enhanced with generative AI at all, according to McKinsey’s “HR Monitor 2025.” Answers are scattered across systems that were never built to talk to each other.

That fragmentation was tolerable when work moved slowly. It isn’t anymore. The same McKinsey report found that 32% of employees don’t have all the skills they need for their current role, and only 12% of HR leaders do strategic workforce planning with a three-year horizon or longer. Meanwhile, the workforce itself is now including AI agents working alongside people.

I believe whoever wins this space won’t win on content libraries or review templates—those are commodities now. It’ll come down to how well a platform handles two very different roles AI has to play.

AI Agents As Actual Teammates​The first role is AI acting as a real member of the team, not a chatbot bolted onto a dashboard. However, in Mercer’s “2024 Global Talent Trends Study,” 74% of executives said they’re “concerned about their talent’s ability to pivot” in this new environment, and only 28% of “HR leaders are very confident they can make human-machine teaming a success.”

That gap between concern and confidence is exactly where a real operating model has to do its work.​ Closing that gap starts with treating an agent’s role definition the way you’d treat a job description: clear scope, clear accountability and permissions tied to the same role-based access controls that already govern human users. If a person on the team needs a login and a defined role, an agent doing similar work needs one too.

Governance isn’t optional here. Gartner projects that by 2030, half of all AI agent deployment failures will trace back to governance platforms that can’t enforce permissions and interoperability at runtime and separately estimates that “over 40% of agentic AI projects will be canceled by the end of 2027, due to escalating costs, unclear business value or inadequate risk controls.” Leaders evaluating any agent-based tool should ask a blunt question before adoption: Can this agent’s permissions be audited the same way a new hire’s would be? If the vendor can’t answer that clearly, the tool isn’t ready for anything beyond a narrow pilot.​

Building Tools On Demand, Not Buying More Of ThemThe second role is different: giving teams the ability to build the exact tool they need inside one system instead of buying another point solution. Every enablement team knows tool sprawl—a tool for surveys, another for content, another for scheduling, another for reporting, none of them talking to each other.

For workforce enablement leaders, that means the next platform evaluation shouldn’t start with a feature checklist. It should start with two questions: Does this system let us govern AI agents with the same rigor we govern people? And does it let us build what we need on one data model instead of stitching together five more subscriptions?

The TakeawayThe convergence of learning, talent and HR operations into one connected function is already underway inside the enterprises I work with, across financial services, healthcare and insurance. What isn’t decided yet is whether AI agents get bolted onto old systems as an afterthought or built in from day one with the same discipline organizations already apply to hiring, managing and governing people. Leaders who start asking their vendors the governance question now—not after the first agent goes sideways—will set the pace for everyone else.

Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?