How Will AI Change the Future of the Workforce and What are the Security Implications?

Dor Sarig

and

March 13, 2024

min read

How will AI change the future of the workforce and what are the security implications? Here's a possible scenario:

Meet Joan, a former real estate agent from the midwest.
She recently landed a junior role in a large tech enterprise, operating a trained LLM-based app that monitors their Security Operations Center (SOC).
Traditionally, this role would need a seasoned security analyst. Because trained models now incorporate expert knowledge, Joan does not need extensive experience to use this technology.

This represents a significant shift in our work paradigm - AI becomes the expert, and employees like Joan manage and supervise its operations.
Since cybersecurity job shortages are extreme and continuing to grow (3.5m in 2023!), many enterprises will have to hire juniors like Joan. This approach is more cost-effective and requires less training.

However, we must not overlook the security implications. As we empower inexperienced operators like Joan, we must ensure security is integral to the model. In unforeseen security or safety events, there should be adequate visibility and control to understand and resolve the issues. And if there is, how can we bridge the knowledge gap and ensure operators understand the model behavior and the implication of its actions?

‍

FAQs

How is AI changing the skills required for cybersecurity roles like SOC analysts?

AI is shifting the skills requirement so that trained LLM-based applications now encode expert knowledge directly into the model. This means employees without extensive security backgrounds can operate sophisticated tools — for example, a junior hire can monitor a Security Operations Center using an AI app that would traditionally require a seasoned analyst.

Why are enterprises hiring junior employees to operate AI-powered security tools?

A cybersecurity talent shortage of 3.5 million unfilled roles in 2023 is forcing enterprises to look beyond traditional hiring pools. AI-powered tools lower the expertise barrier, making it more cost-effective to onboard junior operators who manage and supervise the model rather than requiring deep domain knowledge themselves.

What security risks arise when inexperienced operators manage AI systems in enterprise environments?

When inexperienced operators run AI systems, gaps in visibility and control become critical vulnerabilities. If an unforeseen security or safety event occurs, an operator without deep expertise may not understand the model's behavior or the implications of its actions, making it difficult to identify, contain, or resolve the issue in time.

How can enterprises ensure AI models remain secure when operated by non-expert users?

Security must be built directly into the AI model rather than relying on operator expertise. Enterprises need adequate visibility into model behavior and clear control mechanisms so that even a junior operator can detect anomalies and respond effectively when unforeseen security or safety events arise during production usage.

What is the knowledge gap problem in AI-assisted workforce roles and why does it matter for AI governance?

The knowledge gap problem refers to the disconnect between what an AI model does and what its human operator understands about those actions. As enterprises delegate expert-level tasks to AI systems managed by non-specialists, governance frameworks must bridge this gap — ensuring operators can interpret model behavior and grasp the real-world implications of its outputs.