How much time does AI save at work? New Census Bureau data breaks it down.

Artificial intelligence is increasingly integrated into the workplace, with approximately 55% of U.S. workers reporting its use in their daily tasks. New data from the U.S. Census Bureau indicates that about one-third of these workers find AI significantly boosts their efficiency, cutting down assignment completion time by one to two hours.

Workers are leveraging AI across a variety of functions, including writing and drafting text (32%), generating ideas (32%), and interpreting or summarizing information (31%). Other common applications include searching for information or technical help (37%), administrative tasks (27%), and data analysis or visualization (21%). A smaller percentage use AI for coding, medical care, and logistics management.

The time savings vary, with 25% of AI users reporting less than an hour saved per task. Over 30% see savings of one to two hours, while 15% report gains of three hours and another 15% claim four hours saved. However, the adoption of AI is not always immediately efficient.

Research from MIT highlights that initial AI implementation can decrease productivity, a phenomenon explained by the "J-curve" model. This model describes the temporary dip in output as workers invest time in learning and adapting to new technologies before achieving exponential long-term gains. Kristina McElheran, a University of Toronto professor, notes that AI requires "systemic change" and "introduces friction," particularly for established companies.

“

AI isn't plug-and-play. It requires systemic change, and that process introduces friction, particularly for established firms.