Stanford Launches Real-Time Dashboards Tracking AI's Economic Impact on Jobs

Dashboards created by Stanford HAI’s Digital Economy Lab measure how AI adoption is altering employment trends and job growth, along with broader economic indicators.

Employment in AI-exposed jobs like software development and customer service has dropped by more than 4% this year, while occupations that are reliant on AI automation are showing slowing or shrinking job growth.

Those are some of the key findings from a new set of “dashboards” launched in June 2026 by Stanford HAI’s Digital Economy Lab (DEL). Updated monthly, these dashboards – called AI Economic Indicators – track AI’s economic footprint in real time. Unlike official statistics that report on longer-term trends, these new tools capture swift shifts as they happen.

“At DEL, we’re really interested in understanding the economic impact of AI not only because it has amazing potential, but also because it’s been adopted perhaps more rapidly than any other technology ever,” says Connacher Murphy, who spearheaded the Indicators project. “The issue we’ve kept running into is that it’s very hard to get a big, aggregate view of what’s actually happening, what’s just hype, and what is genuine economic change. Our Indicators can help us get a clearer picture.”

The project pulls data from private companies, academic researchers, and government statistical agencies and organizes it into three primary dashboards that track trends in employment and wages, aggregate outcomes, consumer surplus (the difference between what someone is willing to pay and the actual price they pay), hiring and skills, and AI usage.

One dashboard, the Transformation Tracker, monitors 12 macroeconomic indicators to gauge if any significant changes can be attributed to AI. “We’re asking, ‘Do we see anything happening in aggregate with these bigger, moving statistics?’” Murphy explains.

The third dashboard is the Adoption Monitor, which relies on surveys to track generative AI uptake among individuals and companies. “The baseline goal here is that we want to establish how rapidly this technology is diffusing throughout the economy,” says Murphy.

Dashboard-Derived insights

Using these dashboards, Stanford researchers and their team are keeping tabs on trends in the broader economy, seeing which fluctuate monthly and which have remained stable since launch.

One such trend is early-career divergence, which directly impacts those just entering the workforce. According to the Canaries Dashboard, employment for 22- to 25-year-olds in software development and customer service – the two most AI-exposed sectors – has dropped over 3% annually since ChatGPT’s release in November 2022, with the decline sharpening in recent months to 4.4%.

The Canaries Dashboard also reveals a potential gender disparity, showing slower job growth for young women than their male peers. However, Chandar cautions against reading too deeply into that trend, because ADP Research’s data may not be representative enough to confirm a broader pattern. That said, the preliminary finding aligns with other research indicating that women disproportionately work in AI-exposed fields. At the same time, longer-term government statistics suggest there’s no evidence of a gender gap, Chandar says, before adding, “the gender result is something we want to keep an eye on.”

Another trend the researchers are monitoring is automation versus augmentation. Key data comes from the Anthropic Economic Index, released by the Bay Area-based AI firm, which analyzes roughly a million anonymized user conversations with its large language model, Claude. Tagging users by occupation, the index assesses whether they use the LLM primarily for task automation or task augmentation. “What we find is occupations that really skew toward an automation-heavy pattern have slower employment growth or even contracting employment growth,” says Murphy, especially among early career worker groups.

Looking at the broader macroeconomic picture, the Transformation Tracker shows little movement so far, with its 12 indicators collectively pointing away from explosive growth. For that to change, Murphy suggests keeping an eye on interest rates; currently, 10-year rates remain elevated, offering only “mild” evidence of a surge. “The nice thing about interest rates is they’re fundamentally about expectations of the future,” he says. And because financial markets trade every day, he notes, interest rates offer an earlier glimpse of people’s expectations than any other metric on the Transformation Tracker.

The Stanford researchers continue to expand the project by evaluating new data sources to incorporate, such as AI-driven entrepreneurship. They recently launched the GDP-B: Surplus Observer, a dashboard that tracks the consumer surplus generated by AI. Through online surveys, DEL researchers led by David Nguyen are collecting consumer surplus data across an array of goods, including generative AI tools, coffee, online search, and GLP-1 drugs. “We’re getting a more holistic view of economic value creation from AI systems,” says Murphy.

As data collection across the Indicators continues, the researchers expect interesting trends to emerge. “We don’t want to fly blind into what could potentially be a very large economic transformation,” Chandar says. “While we haven’t gotten definitive answers, we have increasing confidence that we have set ourselves up to catch early indications of transformative economic impact from AI.”

---

Stanford HAI’s mission is to advance AI research, education, policy, and practice to improve the human condition. Learn more.