A Florida State economist just got Microsoft funding to answer a question. Most companies using AI to screen job candidates haven't actually tested it yet.
Does it work better than a human alone?
Pellumb Reshidi, an assistant economics professor at FSU, was named a Senior Fellow of Microsoft's AI Economy Institute this month. He's one of an international cohort studying how AI is reshaping firms and work.
What 'Agentic Hiring' Actually Means
Reshidi's specific focus is agentic hiring. That's the term researchers use for AI systems that screen, interview and evaluate job candidates directly, not just filter resumes by keyword.
That distinction matters. Resume filtering has existed for years. It mostly automates a simple sorting task.
Agentic hiring goes further. It actually conducts parts of the evaluation itself.
He's working with Brian Jabarian of Carnegie Mellon and Luca Henkel of Erasmus University Rotterdam. Microsoft is funding the project through a research grant tied to the fellowship.
"Adoption has moved faster than the evidence," Reshidi said. "Particularly on what happens when an entire industry takes up these systems rather than a single firm, and how the added competition affects firm and worker welfare."
That's a sharper framing than the usual "AI versus humans" debate. Reshidi says the team is approaching the question differently.
"We're approaching them less as a question of AI substituting for people than as one about where the two together outperform either alone," he said.
70,000 Interviews, One Dataset
The team partnered with a major recruiting firm to run a natural field experiment. It covers roughly 70,000 real interviews.
The dataset tracks both recruiters and candidates. It follows them from the initial application through employment and beyond.
That scale lets the researchers study three specific things.
How human recruiters and AI systems interact. How candidates behave differently when they can choose a human interviewer over an AI one. How employers end up competing with each other when they're all relying on similar AI screening tools.
Why This Research Matters Now
This is the third cohort in Microsoft's AI Economy Institute fellowship. It's also the first to study firms directly.
The two previous cohorts focused on education systems. They studied how well those systems prepare people for an AI-driven economy.
Shifting to workplace hiring signals something. It shows where Microsoft thinks the more urgent evidence gap actually sits.
That shift is worth noting on its own. Microsoft has a direct commercial stake in AI adoption succeeding.
Funding research that could find agentic hiring underperforms is a real signal. It shows how seriously the underlying question is being taken.
Findings from Reshidi's team will appear in an edited volume in January. The team is also submitting its research separately to an open-access academic journal.
That second step matters. Publishing in an open-access journal means something concrete.
The actual findings, not just a summary, will be freely available. Any HR team or employer can use them to evaluate these tools honestly.
What This Means for Miami
This is genuinely home-turf research. FSU is a Florida institution.
It's studying a problem Miami employers are already living through. Most are using AI screening tools right now without much solid evidence to guide them.
Miami's hiring managers and HR teams are adopting agentic hiring tools faster than anyone has tested whether those tools actually improve outcomes. Reshidi's research is aimed directly at that gap.
For local businesses, the practical value here isn't abstract.
This research will be one of the first rigorous answers to a question Miami companies are already guessing at.
Does AI screening actually find better candidates? Or does it just find candidates faster?
