Analyst firm Gartner believes that by 2030, all tasks in IT departments will be performed with the help of AI, and there will be no mass job losses.
In a keynote speech at the company's Symposium event in Australia today, vice president analyst Alicia Mullery said 81 per cent of work is currently done by humans acting alone, without the assistance of AI. Gartner believes that in five years, 75% of IT tasks will be human activities powered by AI, and the rest will be performed solely by bots.
Distinguished Vice President Analyst Darryl Plummer said this change means IT will gain workforce and will need to show it's worth keeping. “You never want to look like you have too many people,” he advised, suggesting that technology leaders consult with other colleagues within the company to identify value-add opportunities that IT can implement.
Plummer said Gartner doesn't expect an “AI employment apocalypse” in IT or other industries for at least five years, adding that only 1 percent of today's job losses are due to AI. He and Mallery predicted a decline in entry-level jobs as AI frees up senior staff to take on tasks once reserved for junior employees.
The two analysts also predict that enterprises will struggle to effectively deploy AI as the cost of running AI workloads balloons. Plummer says ERP has a simple up-front cost. You pay to get a license and implement it, and then you pay to train your employees to use it.
AI requires a similar upfront investment, but few organizations can keep up with the pace of innovation from AI vendors. Therefore, introducing AI requires near-continuous exploration of use cases and subsequent retraining.
Plummer said organizations implementing AI should expect to incur 10 unexpected collateral costs, such as the need to acquire new data sets and the cost of managing multiple models.
The need to use one AI model to check the output of other AI models (a necessary step to verify accuracy) is also a cost to consider.
The hidden costs of AI mean that Gartner believes 65% of CIOs are not seeing a return on investment in AI.
Still, Plummer and Mallery suggested that IT executives pursue AI and recommended the big four hyperscalers — AWS, Microsoft, Google and Alibaba — as key suppliers. This is because these companies rival geopolitical superpowers in their ability to mobilize resources and talent. They named OpenAI, Meta, Anthropic, DeepSeek, and XAI as “wildcard vendors” and said all were “not enterprise-ready.”
Plummer said register OpenAI just hasn't done the work to create the licenses that careful enterprise buyers need. Although the suite's Copilot AI component relies on OpenAI technology, the company has done too little to integrate its products with popular enterprise software like Microsoft 365, he said.
They also advised IT executives to quickly retire AI-powered chatbots and replace them with conversational agents with capabilities such as autonomous negotiation with suppliers. ®
