On Monday, Anthropic unveiled the Anthropic Economic Index, a new tool designed to analyze the influence of artificial intelligence (AI) on the labor market and the broader economy. The preliminary findings from the research highlight that software engineering is the sector most significantly affected by this technology. The arts, design, sports, entertainment, and media sectors follow closely in terms of job impact. The report also emphasizes a trend towards the augmentation of tasks rather than outright automation.
The Anthropic Economic Index Shares Initial Report
In a recent announcement, the AI company detailed the launch of the Anthropic Economic Index. This initiative seeks to track the evolving effects of AI on job markets and economic conditions. Distinct from traditional research methodologies, the study leverages data derived from user interactions with Claude, the company’s AI chatbot, rather than relying on expert opinions or general surveys.
The initial report reveals that software engineering roles are leading in AI integration, with 37.2 percent of queries directed to Claude focused on tasks such as software modification, debugging, and network troubleshooting.
Visualized data from the Anthropic Economic Index
Photo Credit: Anthropic
The “arts, design, sports, entertainment, and media” sectors emerged as the second most affected category, accounting for 10.3 percent of interactions. Queries in this area predominantly involved writing and editing tasks. In contrast, roles in “farming, fishing, and forestry” represented only 0.1 percent of the total queries, indicating minimal AI engagement.
The dataset highlights that only four percent of jobs utilize AI for three-quarters of their tasks. Conversely, 36 percent of occupations incorporate AI for at least one-fourth of their responsibilities. The analysis indicates low rates of AI usage in both low-paying and very high-paying jobs, while mid-to-high salary positions tend to adopt AI more significantly.
Interestingly, the report underscores a prevailing trend toward task augmentation—AI collaborating with human users—rather than automation, where AI independently performs tasks. Data indicates that 57 percent of tasks queried through Claude involved augmentation, while 43 percent were attributed to automation.
A closer examination of task performance revealed distinctions between automation—tasks directly performed by AI, such as document formatting—and augmentation—instances where AI supports users. “In just over half of cases, AI was not replacing human tasks but rather assisting in them, engaging in activities like validation, learning, and task iteration,” the report elaborated.
