Australia’s economic engine is sputtering. Labor productivity growth has fallen to a 60-year low, sparking a national conversation about how to reignite growth.
Prime Minister Anthony Albanese is tackling the issue head-on, convening a productivity round table next month. The meeting will coincide with the release of an interim report from the Productivity Commission, which is examining five key areas for reform, including the role of data and digital technologies – and, crucially, artificial intelligence (AI).
The tech and business communities are enthusiastically touting AI as a potential solution. The Business Council of Australia recently declared AI the “single greatest opportunity in a generation” to boost productivity. But does the hype match reality?
What Does Productivity Actually Mean?
Simply put, productivity measures how efficiently we convert inputs – like labor and raw materials – into outputs, or goods and services. Higher productivity generally leads to a better standard of living. Over the past three decades, productivity growth has accounted for 80% of Australia’s income growth.
Productivity can be viewed at different levels: individual, organizational, and national.
Individual productivity is about how effectively a person manages their time and resources. How many emails can you process in an hour? How many products can you inspect for defects in a day?
Organizational productivity reflects how well a company achieves its goals. For a research institution, this might be the number of high-quality research papers produced.
National productivity represents a nation’s overall economic efficiency, often measured as gross domestic product per hour worked. It’s an aggregate of individual and organizational efforts, but tracking how changes at those levels translate into national GDP is notoriously difficult.
AI and Individual Productivity: A Mixed Bag
Early research into the link between AI and individual productivity presents a complex picture.
A 2025 study involving 776 product professionals at Procter & Gamble found that individuals using AI performed as well as a team of two. Similarly, a 2023 study with 750 consultants from Boston Consulting Group showed tasks were completed 18% faster with generative AI.
A 2023 paper examined an early generative AI system used by 5,200 customer support agents at a Fortune 500 software company, revealing a 14% increase in issues resolved per hour. Less experienced agents saw an even larger boost, with productivity increasing by 35%.
However, AI doesn’t always lead to gains.
A survey of 2,500 professionals revealed that generative AI actually increased workload for 77% of respondents. Nearly half (47%) admitted they didn’t know how to unlock AI’s productivity benefits. Barriers included the need to verify AI outputs, the demand for AI training, and unrealistic expectations.
A recent CSIRO study tracked the daily use of Microsoft 365 Copilot by 300 government employees. While most reported productivity benefits, a significant minority (30%) did not. Even those who saw improvements expected greater gains than were realized.
AI and Organizational Productivity: A Difficult Calculation
Attributing changes in an organization’s productivity directly to AI is challenging. Businesses are influenced by numerous social and organizational factors, making it hard to isolate AI’s impact.
The Organization for Economic Co-operation and Development (OECD) estimates the productivity benefits of traditional AI – machine learning applied to specific tasks – to be between zero and 11% at the organizational level.
A 2024 summary paper cites studies showing organizational productivity increases from AI in Germany, Italy, and Taiwan.
Conversely, a 2022 analysis of 300,000 US firms found no significant correlation between AI adoption and productivity, but did find a link with robotics and cloud computing. This suggests AI’s impact may be limited, or simply difficult to disentangle from other technologies.
Sometimes, apparent productivity gains from AI are offset by the need for additional human labor. Consider Amazon’s Just Walk Out technology.
Launched in 2018, the system aimed to reduce labor costs through full automation. However, it reportedly required hiring around 1,000 workers in India for quality control. Amazon has disputed these reports.
More broadly, consider the vast number of people employed to label data for AI models.
AI and National Productivity: The Big Picture
The national-level picture is even less clear.
AI hasn’t yet demonstrably impacted national productivity. It’s reasonable to assume that technology takes time to influence national figures, as companies adapt and build the necessary infrastructure and skills.
However, this isn’t guaranteed. While the internet demonstrably improved productivity, the effects of mobile phones and social media are more debated, with impacts varying across industries, as highlighted in government research.
Productivity Isn’t Just About Speed
The prevailing narrative around AI and productivity focuses on automating tasks and freeing up time for creative work. This is an oversimplified view.
Dealing with your inbox faster doesn’t automatically lead to an afternoon at the beach. More emails simply generate more replies, perpetuating the cycle.
Faster isn’t always better. Sometimes, we need to slow down to be more productive. That’s often when innovative ideas emerge.
Imagine an AI that doesn’t just speed up tasks, but proactively slows us down, creating space for innovation and, ultimately, greater productivity. That’s the real untapped potential of AI.
This article is republished from The Conversation under a Creative Commons license. Read the original article.
