In the first months of 2023, IBM announced it would lay off 8,000 employees to replace up to 30% of repetitive tasks in human resources with artificial intelligence. By taking this step, like Google and Spotify, Microsoft hoped to make employees more productive. IBM’s AskHR, an AI-powered HR assistant, quickly automated 94% of tasks such as leave management, payroll, and employee documentation, delivering $3.5 billion in productivity gains across over 70 business lines.
Still, the results were not what was expected. Instead of cutting jobs, IBM’s number of employees started to increase. Because of AI’s cost savings, CEO Arvind Krishna said IBM could invest more in growth areas, leading to a rise in employment for software engineers, marketers, and salespeople—all roles that require creativity, critical thinking, and emotional intelligence.
As a result, we can see that AI replaces simple jobs but also creates a need for more advanced and human-focused roles. IBM’s example shows that automation can actually help create jobs when companies use the money they save wisely.
Still, the transformation isn’t without hurdles. Support roles continue to be at risk, but those who can handle and deploy AI are highly sought after. Like Duolingo, other companies have found that using chatbots instead of staff often leads to hiring back people when automation fails.
The AskHR case is a good example. In 2024, the company managed 11.5 million interactions and increased customer satisfaction from -35 to +74. Even so, 6% of cases still require people, proving that AI cannot do everything.
IBM’s experience highlights a larger point: AI is changing the job market instead of ending jobs. With the World Economic Forum predicting 92 million job losses to automation by 2030, the challenge now lies in preparing the workforce for roles AI can’t replace yet.
