Dear Class of 2025, Your Job Description Is Already Outdated

First off, congrats. You've made it through years of study, applications, interviews. Now you're stepping into the workforce.

But the workforce you're entering is nothing like the one your professors trained you for.

AI has moved from assistants to autonomous agents to full agent teams. These systems don't just support work. They perform it. Across systems, across tasks, with minimal human input.

Most graduates aren't being prepared for this.

They're stepping into workflows run by AI agents. Agents that make decisions, coordinate execution, and move faster than most onboarding programs acknowledge.

It's not your fault you're unprepared. It's no one's, really.

But if we don't prepare graduates to manage agents, and reset how we onboard and develop them, the responsibility falls on workforce leaders, L&D, and every leader accountable for readiness.

Every enterprise is deploying AI. Almost none can see the work they're deploying it into.

We're still onboarding for a world that no longer exists

Most onboarding still assumes:

• Static job roles
• Linear career progression
• One-size-fits-all training

But work today is structured by tasks and skills. Not job titles.

What the data shows: Microsoft research confirms AI agents already perform complex, high-impact tasks across organizations. Yet new hires still receive outdated materials. Job descriptions, generic training, and no guidance on managing autonomous systems.

What's missing: New talent needs to be equipped with:

• Task literacy
• Prompt fluency
• System design thinking
• Human-AI judgment

These aren't advanced skills. They're the new baseline. Treating AI as an add-on is no longer acceptable.

L&D is not built for this shift

The problem goes deeper than onboarding. The teams responsible for preparing people aren't ready either.

Current state:

• Over 60% of L&D work is administrative and automatable.
• Roles like LMS Admin and Training Coordinator are built around tasks with 80 to 90% automation potential.
• These roles are sunsetting. But the L&D operating model hasn't caught up.

What's emerging instead:

• Capability Architect: Aligns learning with business strategy.
• Skills Intelligence Partner: Translates workforce data into action.
• Learning Experience Designer: Builds personalized, high-impact journeys.

Until L&D retools around these roles and capabilities, it cannot equip new talent. If L&D still manages content libraries, it won't prepare graduates to manage agent teams.

Reimagining onboarding for the AI-ready graduate

Today's early talent enters a workforce shaped by systems, not roles. Onboarding must evolve to reflect that. It's no longer about helping them "settle in." It's about equipping them to operate, audit, and lead in AI-integrated environments. From day one.

Effective onboarding must now:

• Map how work is structured at the task and skill level
• Introduce where and how AI agents operate within those systems
• Clarify where human oversight is essential and how to exercise it
• Build fluency in navigating dynamic, AI-powered workflows
• Embed ethical decision-making into automation design and use

L&D must design onboarding as a living system. Not a one-time event. It connects learning, automation, and performance in real time. Until that happens, new talent walks into complexity unprepared.

The cost of doing nothing

Failing to redesign onboarding means:

• Graduates enter roles they aren't prepared for
• Skills don't match the systems they're expected to manage
• Engagement drops, potential is wasted, and AI investments underdeliver

This isn't a systems issue. It's a talent issue.

If we expect graduates to lead in agent-powered work, we must lead the redesign of how they're prepared. That responsibility sits squarely with leadership, workforce strategy, and L&D.

We can't keep preparing people for outdated work structures and expect them to thrive.

The work has changed.

Graduates know it.

Now leadership must act like it.

Siobhan 💜