Which Jobs Are Most at Risk from AI?
According to a new analysis by OpenAI's chief economist, Ronnie Chatterji, at least 18% of jobs face major AI risk. That means nearly one in five workers could see their roles heavily automated in the near future. But the bigger picture is more nuanced: some jobs will be restructured, while others may actually grow because of AI.
This study looks at over 900 occupations, covering 153.7 million jobs. It offers a fresh way to understand how artificial intelligence will reshape the labor market. Let's break down what this means for you and your career.
The Four Key Questions to Assess AI Impact
The OpenAI framework uses four simple questions to predict how AI will affect a job:
- Can AI do a meaningful share of the work?
- If AI lowers costs, will demand for the service grow enough to keep jobs?
- For tasks AI can't do, is a human still essential for judgment, accountability, or physical work?
- Is AI already being used for these tasks today?
This approach helps separate hype from reality. It avoids two common mistakes: overstating immediate disruption and underestimating long-term change.
Which Jobs Are Most Vulnerable?
The study found that about 25% of jobs have high exposure to AI but also require strong human skills. These roles will likely reorganize rather than disappear. For example:
- Least-elastic occupations (hard to automate): Firefighters, home health aides
- Somewhat-elastic occupations: Physical therapists, editors, dental hygienists
- Most-elastic occupations (high automation potential): Graphic designers, software developers
In jobs with the highest automation risk, AI could perform about 90% of tasks. However, actual usage today is only around 24%. Why the gap?
The Knowledge Debt Problem
Darlene Newman of Ivy Captech Advisors explains that this gap isn't just about technology. It's about knowledge debt. Many organizations have data scattered across spreadsheets, undocumented processes, and rules that only exist in people's heads. When you add AI to such a messy environment, it doesn't magically fix things.
"The model doesn't flag its own uncertainty," Newman warns. "Unlike humans, it pattern-matches against ambiguity. This is why people will remain in the loop longer than most projections suggest."
What This Means for Your Career
If you work in a job with high automation potential, don't panic. Instead, focus on skills AI can't easily replace: creativity, emotional intelligence, complex problem-solving, and physical dexterity. Jobs that require human judgment, accountability, and hands-on work are safer.
For example, while AI can help draft content or generate code, it still needs human oversight for quality, ethics, and context. The future isn't about AI replacing humans—it's about humans working alongside AI.
Final Takeaway
The OpenAI economist admits that predictions aren't exact. Organizational culture, adoption speed, and knowledge gaps will slow down AI's impact. So while 18% of jobs face major AI risk, the transition will take time. Stay adaptable, keep learning, and focus on skills that add unique human value.
Comment