Artificial intelligence companies are increasingly marketing tools designed to automate or augment human labor. Data benchmarks for large language models (LLMs) indicate that these systems, which were previously limited to simple tasks, are now capable of completing more complex software development and financial analysis tasks that typically require up to an hour of human expert labor. Industry reports suggest that investment is shifting toward "AI agents"—virtual workers programmed for specific roles.
Labor market data from Stanford University indicates a 2.7% decrease in employment for 22-to-25-year-olds since the public launch of ChatGPT, with the decline reaching 12.8% in sectors highly exposed to AI, such as software, finance, and creative industries. However, some economists attribute these changes to macroeconomic factors like interest rate fluctuations rather than automation alone. Parallel data from the OECD shows a decline in online job postings for AI-exposed sectors in the United Kingdom, Canada, and the United States.
Corporate adoption of AI has led to a significant increase in "token" usage—the units of data AI uses to process language. While usage for automated tasks has grown, some companies have begun rationing access to these models due to high operational costs. Additionally, there is a growing trend of firms utilizing lower-cost, open-source models as an alternative to more expensive proprietary systems. Experts note that the long-term impact on employment remains uncertain as businesses balance productivity gains against the cost of AI implementation.