For employers and workers alike, the question is no longer if artificial intelligence will reshape work, but how quickly and where. To answer those questions, the Stanford Digital Economy Lab and ADP Research have collaborated to launch the first career-focused measure of AI’s impact on the labor market.

The Canaries Dashboard, built on ADP payroll data, is a near real-time monitoring tool based on “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence,” a Stanford paper that examined AI-exposed occupations and industries for early signals of AI’s effect on wages and the economy.¹

The dashboard is one in a set of AI economic indicators Stanford has built to track and measure the impact of these emerging technologies on the economy.

An early detector

By tracking employment growth by occupation, age, and AI exposure level, the Canaries Dashboard gives us an early indicator system for the labor market.

Putting a microscope to a curated sample of ADP payroll data of 25,000 employers, 4 million workers, and more than 730 occupations, the dashboard can detect changes in hiring and employment in specific occupations and career stages before those effects become visible in aggregated labor-market statistics.

It has already shown us AI’s effects on software development, customer service, and other high-exposure occupations, and its lack of impact on low-exposure work such as home health care.

In the aggregate, AI’s impact on jobs remains modest. Since the introduction of ChatGPT in late 2022, annual employment growth in AI-exposed occupations has increased by 1.1 percent, compared to 2 percent for the least-exposed roles.

When AI’s impact is measured by career stage, however, dramatic differences emerge. For workers aged 22 to 25, employment in highly AI-exposed occupations has been shrinking by 3.8 percent a year, while the least-exposed jobs continue to grow at an annual rate of 2 percent.

Where the canaries sing

AI can automate workflow or augment it. When work is automated, the technology executes individual tasks. When work is augmented, AI acts as a teammate who helps people complete, accelerate, and scale their work.

The Canaries Dashboard also reveals how the type of AI usage affects employment. Occupations primed for AI augmentation have shown more enduring employment growth than occupations where AI is used to automate.

My take

The debate around AI and jobs has been dominated by forecasts and, frankly, guesswork. Will AI replace workers? Augment them? Create new roles?

The reality is more nuanced. AI acts from the bottom up, at the task level, rather than the top down.

The dashboard captures this task-level change to reveal where AI’s true value – and its risk – lie. In occupations and career stages where AI amplifies human abilities and potential, we see employment growth.

With this value in mind, the Canaries Dashboard can help employers, workers, and policymakers move from guesswork to informed decision-making to shape the future of work, one task, one occupation, and one career stage at a time.

Note: Some readers last week noticed that our National Employment Report data history file didn’t match what we’d been reporting in NER Pulse. The file we released was based on an outdated reference file. We’ve updated that file and the accurate data can be found here. Thanks for the catch!

Main Street Macro will be on hiatus next week. The NER Pulse will be released as scheduled on June 23.

Download this week's NER Pulse data