OpenAI has outlined the areas of work where human involvement is likely to remain important even as artificial intelligence becomes capable of performing a growing share of workplace tasks, highlighting accountability, human relationships and physical presence as key factors.
The findings come from OpenAI's AI Jobs Transition Framework, which examines more than 900 occupations covering 99.7% of US employment. Rather than attempting to predict which jobs AI will eliminate, the research evaluates how occupations could change as AI capabilities expand.
A central finding of the research is that technical exposure to AI does not automatically translate into job displacement. OpenAI's framework combines AI exposure with what it calls "human necessity", alongside demand elasticity and observed ChatGPT usage, to assess where labour-market changes could emerge.
Human necessity captures situations in which workers remain central to delivering a service even when AI can perform some of the underlying tasks. OpenAI identifies three broad forms: regulatory or accountability requirements, relational interactions and physical presence.
Regulatory and accountability requirements apply where laws, professional standards or institutional processes require a person to remain responsible for decisions or outcomes. This can affect professions where licensing, approval and accountability remain attached to human practitioners.
Relational necessity covers work where trust, care and direct human interaction are important to how a service is delivered. OpenAI points to professions such as teaching and nursing as examples where AI may assist with tasks while interpersonal interaction continues to play a significant role.
Physical necessity applies to jobs requiring people to perform work in real-world environments. Tasks involving physical execution can remain dependent on workers even as AI assists with information processing, planning or other cognitive components.
The report also considers demand elasticity, or whether making a service cheaper through AI could increase demand enough to offset productivity gains. This means AI could affect occupations differently even when their technical exposure to the technology appears similar.
OpenAI cautions that its framework is intended to examine near-term labour-market pressures rather than provide definitive predictions about individual jobs. The company also notes that technological capabilities and the boundaries between occupations can change over time.
The findings come as workplace AI use expands beyond existing job descriptions. Separate OpenAI research analysing more than 1.5 million work-related ChatGPT messages found workers repeatedly using AI for tasks traditionally associated with other occupations, suggesting that job responsibilities may broaden even when job titles remain unchanged.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.