Join our daily and weekly newsletter for the latest updates and exclusive content on industry-leading AI coverage. learn more
AI is reshaping modern workplaces, but up until now it has been difficult to quantify the impact on individual tasks and occupations. The AI startup behind Claude, a new report on humanity, offers a data-driven view of how businesses and experts integrate AI into their work.
Released today, the Human Economic Index provides a detailed analysis of industry-wide AI usage, drawn from millions of anonymized conversations with human AI assistant Claude. The report shows that AI has not yet widely automated the entire job, but is widely used to enhance specific tasks, particularly in software development, technical writing and business analytics.
“The use of AI is primarily focused on software development and writing tasks, which considers almost half of the overall usage,” the report said. “However, AI use has expanded more widely across the economy, with about 36% of occupations expanding using AI in at least a quarter of the related tasks.”
Not only hype: Humanity offers a ground-level view of AI adoption
Unlike previous studies that relied on expert predictions and self-report investigations, human studies are based on a direct analysis of how workers actually use AI. The company leveraged its privacy analytics tool Clio to explore more than 4 million user conversations with Claude. These interactions were mapped to occupational categories in the U.S. Department of Labor’s O*net database.
Data suggests that AI plays an important role as a collaborative tool rather than simply acting as an automation engine. In fact, 57% of AI usage in the dataset were related to “enhancement.” In other words, AI supported workers rather than exchanged. This includes tasks like brainstorming, refine your ideas, and verifying accuracy. The remaining 43% of use were categorized into the direct automation category, with AI performing tasks with minimal human involvement.
This balance between scaling and automation is an important indicator of how today’s businesses are deploying AI. “We see that 57% of interactions exhibit enhancement patterns (before and after the task) and 43% suggest automation (satisfaction of requests with minimal human involvement),” the report states, It’s there.
Partners over exchange: AI is increasing rather than eliminating jobs
One of the most prominent conclusions of the report is that AI has not abolished the role of the entire job. Instead, they are being selected to assist with specific tasks rather than fully automate the occupation.
“Only about 4% of occupations show AI use in at least 75% of tasks, suggesting the possibility of deep task-level use in several roles,” the report states. “More broadly, it shows that 36% of occupations are used in at least 25% of tasks, indicating that AI is already beginning to spread into the task portfolio in a significant portion of the workforce. ”
This selective recruitment suggests that while AI is converting jobs, it has not yet led to widespread work movements. Instead, experts use AI to increase productivity, enhance repetitive tasks, and improve decision-making.
The report identifies software engineering as the area where AI adoption is the highest, accounting for 37.2% of conversations analyzed. These interactions typically included tasks such as debugging code, changing software, and troubleshooting the network.
The second highest category of use was in creative and editing work, including the roles of media, marketing and content production (10.3% of queries). AI is widely used to help draft and refine texts, research and generate ideas.
However, AI use was significantly lower in areas that require physical labor, such as medical, transportation, and agriculture. For example, only 0.1% of the conversations analyzed were related to agriculture, fishing and forestry challenges.
This disparity highlights the current limitations of AI that excel in text-based and analytical tasks, but struggle with work that requires practical work, manual dexterity, or complex interpersonal interactions. It’s there.
AI wage gap: Amazing sweet spots for adoption
One of the most interesting findings in the report is that when AI usage correlates with wages, it does not follow a simple pattern. Rather than focusing on low- or high-paying jobs, AI adoptions peak in the medium- to high-paying range.
“AI uses peaks in the top quartiles of wages, but falls at both extremes of wage range,” the report states. “Most high-rise occupations gathered in the upper quartiles primarily cater to positions in the software industry, but both very high-paying occupations (doctors) and low-paying positions (restaurant workers) are relatively high-paying occupations. It shows low usage.”
This means that AI is most actively recruited in roles that require analytical and technical skills, but not necessarily the highest level of professional expertise. It also raises important questions about whether AI exacerbates or mitigates existing economic inequality. Especially when low-wage workers have less access to profits that increase productivity in AI.
What business leaders who know AI will form the workforce should know
For technical decision makers, this report provides a roadmap where AI is likely to have the greatest short-term impact. The data suggest that companies should focus on AI adoption in knowledge-based occupations where enhancement is the dominant pattern rather than complete replacement.
The report also provides early warnings to policymakers. Although AI has not yet replaced the entire large-scale work, an increase in presence in high-value tasks can have a significant impact on workforce dynamics.
“AI is already beginning to spread across a significant portion of the workforce into the task portfolio,” the report states. “Our data reveals where AI is used today, but we infer long-term outcomes from these early usage trends.
Humanity has opened sourced the dataset behind its analysis, and invited researchers to explore further how AI shapes the economy.
The AI economy is here – Are you ready?
Human economic indexes provide one of the most comprehensive snapshots of how AI is used in the workplace rather than theoretical. The findings suggest that AI has not led to the displacement of the masses’ work that many fear. It changes the nature of work in meaningful ways.
For businesses, this means that adopting AI not only reduces costs, but also unlocks new efficiency and creativity. For policymakers, it raises urgent questions about how to ensure that AI benefits are distributed fairly, rather than deepening existing economic divisions.
The challenge for the future lies not only in measuring these changes, but preparing them. As AI continues to expand its role in the workforce, businesses and workers will thrive, learning how to use it effectively. Those who ignore it are at risk of being left behind.
