IBM is planning to use artificial intelligence and automation to change how some back-office work gets done. CEO Arvind Krishna told Bloomberg News that the company is freezing hiring for positions that could be replaced in the coming years, with human resources among the areas in focus.
Where IBM expects automation to have an effect
The company’s estimate centers on roughly 26,000 back-office positions. About 30 percent of those roles could be replaced by AI and automation within five years, a figure that amounts to about 7,800 positions.
Krishna said he can “easily see” that number. The estimate describes potential role reductions, but not every position would necessarily disappear through a direct layoff. Some of the change is expected to come from natural attrition and from leaving positions vacant when employees depart.
A hiring freeze for roles considered automatable is one way IBM could manage that shift. As employees leave, the company could choose not to refill certain positions, while automation takes on some of the work those roles handled.
Routine HR work may change first
Some human resources tasks are straightforward enough to be fully automated, according to Krishna. Examples include issuing work certificates and transferring employees. These are administrative processes that can be handled through repeatable steps.
Other HR responsibilities involve judgment about people and organizations. Krishna said assessing workforce structure and productivity is unlikely to be replaced over the next decade. That distinction suggests IBM sees AI as more suited to routine processing than to every decision involved in managing a workforce.
For employees, the practical impact may therefore vary by task. A role that includes repetitive administration could change substantially even if work requiring evaluation and organizational understanding remains with people.
Hiring and cuts can happen at the same time
IBM employs about 260,000 people and says it is continuing to hire for software development and customer-facing roles. In the first quarter, the company added 7,000 employees, even as it had announced plans to cut about 5,000 jobs.
Those figures show that the company’s workforce plans are not a single across-the-board reduction. IBM is describing different trajectories for different kinds of work: some positions may not be refilled as automation expands, while hiring continues in software development and roles serving customers.
The company also expects productivity and efficiency gains to save $2 billion annually by the end of 2024. The source does not specify how much of that projected saving would come from AI, so it should be read as a broader efficiency target rather than a direct measure of AI-driven job replacement.
Why generative AI has raised the stakes
GPT-4 has prompted managers to consider whether AI could do more than generate text. The source describes it as a potential interface for making individual workers more productive and for automating complex processes with tools such as Auto-GPT.
That prospect has fueled both concern and optimism about work involving data, text, and code. The issue is not limited to whether a job title disappears; AI could also alter the tasks within a job, shifting which parts are performed by people and which can be handled by software.
OpenAI recently published a study on potential labor-market effects. It estimated that large language models could affect at least ten percent of the jobs of about 80 percent of U.S. workers, and affect at least 50 percent of tasks for 19 percent of workers. These estimates point to broad task exposure, while IBM’s plans illustrate how one company is approaching a narrower set of positions.
For now, IBM’s stated approach combines a hiring freeze in roles it believes could be automated with continued recruitment in other areas. The scale of change will depend on which tasks can be handled reliably by AI and automation, and on how the company fills or leaves open positions as its workforce evolves.