Economists and business leaders are increasingly debating whether new productivity data shows the first measurable signs of AI-driven economic impact, as organizations begin moving from early experimentation toward redesigning workflows around AI systems. Recent analysis cited by Erik Brynjolfsson, director of the Digital Economy Lab at Stanford University, suggests U.S. productivity may have risen significantly in 2025 even as job growth slowed.

The discussion matters because productivity has been one of the missing indicators of AI’s real-world value. While AI adoption has accelerated, many economists argued that macroeconomic data still showed limited evidence that AI was improving efficiency across industries.

One explanation is the “J-curve” effect, where general-purpose technologies require major investment and organizational adaptation before measurable gains emerge. In this model, businesses must redesign workflows and integrate systems before improvements appear in economic output.

The debate affects enterprise leaders, operations teams, economists, and organizations deciding whether AI should remain a supporting tool or become a core operational layer that reshapes how work gets done.

Here’s what the emerging productivity debate could mean for companies transitioning from AI experimentation to structural workflow redesign.

The discussion gained attention after economist Erik Brynjolfsson, director of the Stanford University Digital Economy Lab, argued that revised U.S. labor data may signal rising productivity.

Brynjolfsson, who has studied AI’s economic effects for years, previously published research suggesting that AI may disproportionately impact entry-level workers — particularly those ages 22 to 25 in highly AI-exposed roles — underscoring how productivity gains and workforce shifts may unfold simultaneously.

According to the summary published by Yahoo Finance, job growth estimates for 2025 were revised down to 181,000 — from an initial reading of 584,000 and well below 2024’s gain of 1.46 million — even as GDP growth remained strong. With fourth-quarter GDP tracking up 3.7% despite slower hiring, analysts argue that output continued to expand with fewer workers, suggesting a possible surge in productivity.

Brynjolfsson’s analysis estimates that U.S. productivity increased roughly 2.7% in 2025 — nearly double the 1.4% annual average seen over the previous decade. He argues that this could signal a transition from an AI investment period into a “harvest phase,” where earlier efforts begin producing measurable output.

The idea follows a historical pattern often seen with general-purpose technologies: early investments obscure gains until organizations begin restructuring work around the technology.

Not everyone agrees that AI’s economic impact is already clear.

Torsten Slok, chief economist at Apollo Global Management, noted that employment, inflation, and productivity measures still do not consistently show widespread AI effects. The caution echoes earlier debates around past technological revolutions, where visible productivity gains lagged behind adoption.

Others see early indicators pointing in the opposite direction. Stephen Brown, chief deputy North America economist at Capital Economics, argued that information and communication technology (ICT) output in technology industries rose in the third quarter even as employment declined, suggesting productivity improvements could already be emerging as AI adoption expands.

Even Brynjolfsson cautions that additional periods of sustained growth are needed before declaring a long-term productivity trend, noting that outside economic forces could affect outcomes.

This uncertainty highlights a key tension: early data may signal change, but confirmation requires consistent results over time.

Beyond macro data, the most significant insight may be how organizations are using AI internally.

Erik Brynjolfsson describes a divide between companies using AI for limited tasks and a smaller group of “power users” automating end-to-end workflows with AI agents, completing work in hours instead of weeks.

As Brynjolfsson wrote in the Financial Times, “We are transitioning from an era of AI experimentation to one of structural utility. We must now focus on understanding its precise mechanics. The productivity revival is not just an indicator of the power of AI. It is a wake-up call to focus on the coming economic transformation.”

This distinction helps explain why productivity gains may appear uneven. If only a small number of organizations has redesigned processes around AI, macro-level improvements would likely emerge slowly and inconsistently.

The implication is clear: the biggest productivity gains may come not from AI tools themselves, but from organizational willingness to redesign workflows around them.

Q: What is the J-curve effect in AI productivity?
A: It describes how major technologies often require upfront investment and workflow changes before measurable productivity gains appear.

Q: Why are some economists skeptical about AI’s impact?
A: Macro indicators such as employment and inflation have not yet shown consistent, widespread evidence of AI-driven improvement.

Q: What separates AI power users from other companies?
A: They automate complete workflows using AI agents rather than applying AI only to individual tasks or assistance features.

Q: Does the data prove AI is transforming the economy?
A: No. Analysts say more sustained growth periods are needed before confirming a structural trend.

Q: What should organizations focus on if productivity gains depend on workflow redesign?
A: Organizations may need to move beyond isolated AI tools and evaluate entire workflows where automation or agent-driven processes can reduce process friction and time-to-output.

These discussions suggest that the next phase of AI adoption may be less about the capabilities of new models and more about whether organizations restructure work to take advantage of them.

Who should care: Enterprise leaders, operations teams, strategy executives, and organizations deciding whether AI pilots should evolve into broader transformation initiatives.

Why it matters now: Early productivity improvements may reflect organizations that are redesigning workflows rather than simply adding AI assistants to existing processes. This suggests competitive advantages may emerge from operational change rather than technology access alone.

What decision this affects: Companies must decide whether AI remains a productivity enhancement within existing workflows or becomes a catalyst for redesigning how work moves through teams, systems, and decision chains.

Organizations that treat AI as a layer added on top of legacy processes may see incremental gains, while those willing to rethink workflows end-to-end could unlock larger productivity improvements. As AI adoption matures, the organizations that gain the biggest advantage may not be those experimenting with the most tools — but those redesigning work itself to turn AI from an assistant into structural utility.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.