Overview
Artificial intelligence is already delivering pronounced productivity increases inside specific, well-defined company activities, but those localized advances have not yet translated into a major boost to overall economic output, according to a report published by BofA Global Research on Friday.
Current measured impact
The research synthesis shows that, at present, the integration of AI into business processes contributes roughly 0.1% per year to aggregate macroeconomic productivity. The gains are concentrated in narrow tasks where model application and workflow redesign are straightforward, producing clear efficiency improvements at the unit or process level.
Why macro effects remain limited
BofA Securities economists point to several practical headwinds that limit the near-term translation of those task-level productivity gains into broader, economy-wide output. The report cites delayed corporate adoption timelines - companies may take months or years to scale pilot projects across operations - persistent skills gaps among workers needed to deploy and operate AI tools, and institutional organizational constraints that complicate reconfiguring processes and management structures.
In addition, the report describes its GDP tracking mechanics: the tracking calculation mechanically aggregates high-frequency data releases to align with official government tracking metrics, a methodology used to assess recent output trends alongside AI impacts.
Potential upside over time
Longer-term, the analysts allow for substantially larger structural effects as AI models iterate and expand their capabilities. Under a baseline scenario in which models broaden the range of tasks they can perform, penetrate more industry sectors, and fall in cost, the report estimates macroeconomic productivity gains could increase by a factor of ten - reaching as much as 1.0% per year over the next decade.
In that optimistic baseline, the acceleration in efficiency could push long-term global GDP growth to an annualized pace of up to 4.5%.
Concurrent indicators and policy context
The BofA team also flagged that U.S. second-quarter GDP tracking remained steady at a 2.6% quarter-over-quarter seasonally adjusted annual rate (saar) after the release of April housing starts data. The broader economic calendar on Friday included remarks from Federal Reserve Governor Christopher Waller on the domestic outlook, which coincided with the official swearing-in ceremony of Kevin Warsh as the new Federal Reserve Chair.
On the consumer front, final May University of Michigan Consumer Sentiment printed at 48.2, a slight beat versus consensus expectations that had penciled in a small decline to 48.0 for the period.
Bottom line: AI is delivering visible productivity benefits in targeted corporate applications today, but BofA's analysis underscores that meaningful macroeconomic uplift will depend on faster and broader adoption, workforce readiness, and organizational change. The scale of potential gains is material in the longer run, but the pathway to that outcome is constrained by practical frictions in the near term.