AI Helps and Hinders, but Employees Can Get the Most Out of It With These 6 Skills

AI helps employees do their jobs: On average, they spend an hour and a half each day using AI.

But it’s not all good productivity: Employees spend about half an hour of that time troubleshooting AI errors and on prompt iteration, according to BambooHR’s report, Redesigning Work: AI’s Performance Review.

This hasn’t stopped employees’ enthusiasm for AI: 65% are confident and excited about using it, mostly fueled by expectations that they’ll save time. And many companies use it as the organizational knowledge base: More than a third of employees say knowledge is transferred primarily through AI tools rather than humans.

AI Helps Build Enthusiasm, Productivity

Enthusiasm can’t replace actual productivity, though. And productive knowledge transfers can’t happen if employees don’t have the right skills to use AI properly.

While every organization is different and AI use can be specific to roles and outcomes, there are some common, essential skills for employees to get the most out of AI.

Gartner has practice identified six must-train-for skills to help employees manage AI, rather than become overwhelmed by its capabilities and their expectations to master the evolving tech.

“Organizations are treating AI skills as an ever-expanding checklist,” explained CV Viverito, a Senior Director Analyst at Gartner explained in the Harvard Business Review. “A better approach focuses on developing fewer core skills that have the potential to evolve as AI also evolves.”

The six skills fall under two core areas: work-orchestration skills and oversight and governance skills. Here’s what Viverito and Gartner shared.

Work-Orchestration Skills

These are skills employees likely already have or have experimented with: how they work with AI to get results. More specifically, you’ll want to help them improve:

Problem Formulation

Employees want to build problem-formulation skills. That means they take the right problem, with the right background information and ultimate goal to AI.

AI tool’s output is only as good as the input it receives. So employees need to lead work with their knowledge rather than letting the technology lead.

Many organizations are moving away from generic or self-taught AI training to workshops organized by HR and IT that scope and define business challenges a problem statement, determine objectives, and establish boundaries to guide effective solutions.

You can find HR-specific prompts here: 25+ Useful Gen AI Prompts for HR Pros. More importantly, there’s a formula for creating specific, clear prompts.

Systems Thinking

According to Viverito, systems thinking is the ability to understand how individual components interact and connect within the larger whole. It’s important because systems thinking improves employee performance by 1.6x where AI is being used in the workplace.

The reason is relatively obvious: When employees work thinking about the system, they’re out of their silos and overcoming complexity becomes easier.

But how? You want to create formal ways for employees to understand the ripple effects of an interconnected system. For instance, Jean in receiving needs to understand how her workflow affects accounts payable. Input mistakes, bottlenecks or failed AI oversight can lead to billing issues.

Some companies are mapping what front-line employees do before introducing company-wide AI systems so they can bear out real workforce barriers, tedious workarounds and other bottlenecks. This also helps gain employee buy-in to systems thinking.

Strategic Foresight

Once organizations move employees through problem formulation and systems thinking, the next step is to strategic foresight, creating new ways of working that improve scale and possibly disrupt markets.

Employers will want to help employees shift from just using AI as a productivity tool to using it to anticipate future scenarios and building adaptive solutions. 

“The goal is not to successfully predict the future, but to build change-ready mindsets that anticipate disruption,” Viverito wrote.

The best tool is actually an old tool: Embed “what-if” scenario-planning into existing decision-making processes and AI processes as they evolve, prompting employees to anticipate and plan for the best, worst or any disruption.

Oversight and Governance Skills

Oversight and governance skills are the ways employees ensure AI results are accurate, safe and trustworthy for your organization to use. Those include:

AI Judgment and Risk Mitigation

The researcher noted that inherent trust in AI outcomes and overreliance on the tools leads to errors, biases and compliance issues. So beyond using and mastering AI, employees will need to verify reliability, mitigate risks and practice ethical AI governance.

That’s more than just “a human in the loop.” It starts with formal training and guiding principles on responsible AI use. Beyond that, teams will need to develop shared expectations for questioning and checking AI’s work.

The researcher shared one success story along these lines: A marketing research firm developed a verification audit process. Employees are required to be fact-checkers for their first few AI-generated briefs. They follow a structured scorecard to verify data, cross-referencing back to original documents and websites. They log unvalidated claims to help fine-tune future research.

Human-AI Mediation

“Employees will increasingly need to interpret ambiguous situations and make nuanced decisions in an AI-enabled work environment,” Viverito wrote.

AI changes how we work nearly every day. Automation will likely increase, so human judgment will meet machine logic more often. And employees will need to override or comply based on human logic.

The author used this example: AI cancels a customer’s subscription after a few missed payments, but it doesn’t recognize that the customer was impacted by a natural disaster — something a human would make an exception for.

One solution is having teams do post-mortems and even pre-mortems (similar to the what-if scenarios mentioned above) to review and anticipate how to handle complicated exceptions to the rules AI follows.

Ethical Leadership

Verifying outputs and ensuring that AI delivers fair and correct results isn’t just leadership’s responsibility. Every employee who uses AI will need to have the ability to have a hand in managing the design and operation of the AI systems you employ. It’s everyone’s responsibility to shape a trustworthy system.

According to Viverito, one tactic is “red teaming.” Employees systematically and intentionally test their AI systems to find weaknesses. They act as internal ethical hackers, trying to identify biases or purposely confuse autonomous workflows before they’re rolled out to a larger population.

“Ultimately, building workforce skills that evolve in lockstep with the maturity journey of the technology itself is the input needed to drive real AI value,” Viverito wrote.