As artificial intelligence advances from automation to autonomous agency, the workforce required to defend our digital infrastructure is undergoing a fundamental shift. Earlier this month, the AI Workforce Consortium, a coalition co-founded by Indeed alongside technology and workforce leaders including Cisco, Accenture, Google, IBM, Microsoft, and Intel, released its newest research, AI Agents and the Impact on Cybersecurity.

Formed in 2024, the AI Workforce Consortium brings industry leaders together to track job transformation, identify emerging skills gaps,  and provide data-driven strategies for workforce readiness.  As a founding member, Indeed contributes real-time labor market insights to help business leaders and policymakers navigate these shifts.

The first in a new Spotlight series examines how agentic AI is reshaping cybersecurity roles, skills, and hiring. Combining original findings with Indeed global jobs data, the research points to a clear conclusion: solving the talent shortage requires more than technical upskilling. It requires rebuilding entry-level career pathways, establishing standards for human oversight, and modernizing public policy for an AI-driven economy.

Cybersecurity is our first focus area because agentic AI adoption is furthest along and where the resulting skills gap is most defined. As cyber threats evolve, the workforce defending against them must evolve just as fast.

Key Findings: 

  • Global talent shortage: 4.8 million cybersecurity roles remain unfilled worldwide, even as demand grew 9.5% in G7 countries over the last six months.
  • The AI surge: Postings requiring AI cybersecurity skills have doubled year-over-year.
  • Skill shifts: Demand for ethical reasoning (+533%), systems thinking (+251%), and stakeholder engagement (+125%) is spiking alongside technical skills.
  • The entry-level bottleneck: Senior hiring grew 65%, while junior roles grew just 5.9%. Employers report that entry-level candidates lack hands-on experience with AI agents (49%), technical depth (48%), and human judgment (45%).
  • From processing to oversight: As AI agents handle repetitive work like alert triaging, analyst roles are shifting from manual data processing to high-level strategic oversight.

What This Means for Employers & Workers

As AI agents absorb more routine technical execution, employers are placing greater value on human judgment: validating outputs, assessing risk, and determining when a system’s output should be questioned rather than accepted. Yet, the data reveals a widening gap: companies want experienced talent, but early-career workers lack access to the AI tools needed to build that experience before entering the workforce.

Indeed’s own data reflects the same pattern across industries. Closing this gap means employers recruiting strategically, upskilling the workforce already in place, and rebuilding entry-level pathways around experiential learning, apprenticeships, and industry-sponsored projects.

Recommendations for Policy & Industry Leaders

Preparing workers for the AI era isn’t something one company or school can do alone. Government, business, and education must work together to keep our digital world safe and help workers adapt. Leaders across every sector should focus on four priorities:

  • Leverage real-time, accessible workforce data:  The right kind of data is essential for effective policymaking, impactful workforce development, and strategic economic planning. Official government labor statistics move too slowly. Policymakers should modernize labor market datasets, and public-private partnerships should be prioritized, to provide real-time visibility into which skills are rising and where hiring demand is shifting.
  • Build on-the-job AI training: AI upskilling shouldn’t be an after-hours burden for employees. Employers, industry groups, and training providers need to build structured learning directly into normal career growth, backed by credentials recognized across the whole sector.
  • Expand portable credentials across industries: Prioritize workforce programs to include short-term, refreshable certifications, alongside longer traditional training, recognized across industries. Qualifications travel with workers across employers, regions, and borders without forcing them to re-prove basic skills.
  • Establish clear standards for human oversight: As AI agents handle more frontline security work, human roles can shift from manual tasks to oversight. Policymakers and industry leaders should  further develop  what “meaningful oversight” actually means in practice,so cross-border organizations have consistent standards to follow.

AI-driven threats will keep evolving and therefore so must the workforce built to counter them. Defending digital infrastructure requires skilling systems grounded in real-time labor data, hands-on tool access, and shared standards for human oversight.

Want to learn more? Check out the next AI Workforce Consortium Webinar on October 12 – register here.