The rise of agentic AI could help Australia address its low productivity growth problem while overcoming a burnout epidemic in the country, according to intelligent automation technology provider UiPath.
Australia’s productivity growth averaged just 1.2% throughout the 2010s. However, Australia’s Federal Treasurer, Jim Chalmers, noted that the Federal Treasury has downgraded its long-run forecasts for productivity from 1.5% to 1.2%. He has said “there is no more important structural problem” to fix in the economy.
Burnout may be a factor impacting workplace productivity. A survey of 1,000 full-time office workers by recruiter Robert Half found that 80% of Australians were experiencing some level of burnout in 2024. These findings are supported by Microsoft’s 2022 Work Trend Index, which revealed that 62% of Australian workers were burned out, higher than the global average of 48%.
Yelena Galstian, UiPath’s local head of solutions and customer advisory, told TechRepublic that leveraging AI and automation could help organisations attain productivity gains. It could also help individuals reduce burnout by providing more balance through technology support for manual tasks.
Agentic AI will be able to make decisions and perform actions
Agentic AI is viewed as the next step in automation. IBM combines planning, memory, tool usage, and autonomous actions to perform various workplace tasks. IBM envisions a future with armies of agents that can interact with humans and perform autonomous actions.
Galstian said agentic AI, named one of Gartner’s top 10 trends this year, enables automation to expand into decision-making tasks. AI agents will no longer rely on humans to outline problem-solving steps. They can reason independently and interact with tools, such as robotic automation systems, to perform actions.
AI agents poised to handle more cognitive steps
Previously, process automation technologies were leveraged to ease the burden of workers completing rules-based tasks, such as data extraction and entry. Galstian noted how the invoice dispute resolution process allowed some tasks to be performed by automation to cut down on time.
“Anything that would have been about reading and processing the invoice and entering it into systems, anything that would have been about a quick reply, or entering the result into the final finance systems; all of those types of things are classic automation skills,” Galstian explained.
However, as these agents become more advanced in the coming years, they can handle more cognitive steps in a process. They will determine appropriate resolutions independently, involving humans only for escalations, clarifications, or when specific criteria require human oversight.
SEE: How Generative AI Is Becoming A Source of ‘Costly Mistakes’ for Enterprise Tech Buyers
“The amount of time a human agent will need to spend on invoice dispute resolution will be significantly less than it would have been with traditional automation,” Galstian said.
“With agents, we’re giving up that curve of reliability for agency; we’re giving them the opportunity to act independently and make dynamic decisions. AI agents are good for non-deterministic type scenarios … where I might be giving the same input, but there are multiple ways a problem can be solved.”