Companies are cutting jobs while betting on AI gains — real or anticipated. Atlassian cut 10% of its workforce, or about 1,600 employees, to fund increased investment in AI development, while Block slashed roughly 4,000 of its 10,000 employees, a move the fintech tied to AI’s potential to automate work.
Such layoffs suggest that company leadership is acting on expectations that AI will automate thousands of jobs. As more companies pursue headcount reductions to boost efficiency – and are rewarded for it by investors — the message becomes clear: AI is not just being positioned to augment human work.
AI washing – using AI as a convenient explanation for decisions like layoffs – may be giving companies cover, using the buzz around AI to mask the need to cut costs or earlier overhiring. But the idea that AI is already replacing jobs doesn’t line up with reports that productivity gains from AI have so far been underwhelming.
Low-quality or inaccurate AI-generated content – dubbed AI “work slop” — may be part of the reason productivity gains have lagged. Such content can look useful, but upon closer inspection, it often creates more work to review, correct or clean up.
Managing the digital mess
Despite the mess AI might spawn, it offers CIOs an opportunity to demonstrate their leadership, said Sumit Johar, CIO at finance software company BlackLine, noting “CIOs are in the best position to drive the transformation within their organization. But managing the AI mess, he stressed, requires strategic planning that accounts for the capabilities and limitations of AI.
For example, are the enterprise’s AI capabilities advanced enough to make entire roles redundant? Or are the proposed cuts intended to free up money for AI investment that will hypothetically take over the work of the people cut loose?
“Everybody’s convinced there is potential in AI to drive dramatically higher employee productivity, Johar said, which, combined with automation, would offset job cuts. But the disconnect seems to be how soon that can happen.”
For the moment, he remains skeptical of the speed at which businesses will be able to hand over entire processes to independent AI systems.
“At least in … the circle of companies and CIOs I speak to, people are being very measured about handing over the keys of any business process to AI for end-to-end autonomy,” he said.
Some cuts might well serve a company’s bottom line and investors, but premature, overenthusiastic cuts might come back to haunt CIOs, warned Shelley Seewald, CIO at Tungsten Automation, an automation software company.
“We do have companies that are rehiring some of the folks that they’ve let go because maybe they didn’t get the outcomes they were expecting,” she said.
The work slop problem
Meanwhile, AI work slop can gum up the works rather than help teams work more efficiently. “That’s definitely the one thing every company sooner or later has,” said Seewald.
CIOs must learn to recognize slop and figure out what that means for their teams and for their enterprise’s outcomes.
AI tools are designed to give users an answer, even if that answer ends up being unhelpful or wrong. When not enough upfront work goes into training, monitoring, and governance, the risk of churning out slop increases – creating more work for employees.
“The more you know your topic, the easier it is to spot the slop,” said Seewald. Well-trained eyes might call it out, but CIOs need formalized processes to measure the actual value of AI.
Johar also underscored the importance of formal processes: “How do you measure yourself? How do you compare against others?” Every organization that wants to leverage AI must ask these questions, he said, but it requires rigorous assessment and benchmarking.
At Blackline, surveys across different departments gather employee feedback to garner information beyond engagement. “Every quarter we’re asking our employees, how is it helping? Where is it helping? Are you running into any other challenges? Is it really making you productive? How much time it is saving for you?” Johar shared.
Paying attention to employee morale
In a not-too-distant future, CIOs may be managing more AI agents than human employees. But there is still a human workforce that needs leadership today.
“If you create a situation where employees have to constantly be scared about, ‘Am I next to basically lose my job,’ you cannot build the culture of transformation within the company,” said Johar. “People should not be scared of transformation.”
He argued that CIOs must view AI adoption as more than technological transformation. “That’s not the way you’re going to win this transformation battle,” he said. “It needs to be a culture-, people-focused transformation.”
And people are getting burnt out. “I think the biggest concern I have, and I think a lot of people are starting to see it, is the AI burnout,” said Seewald.
That burnout can further stall productivity, and CIOs have to consider how to balance AI use and employee capacity.
Managing the talent pipeline
The velocity and volume of demands on AI make it difficult for CIOs to plan long-term. But leaning too hard into the here and now can be short-sighted. Seewald is hearing a lot of excitement about AI agents and the need for less human talent.
“Talking to my peers, there is this, ‘Oh well, we just need those senior-level roles that can provide some oversight over those agents,’” she said. “But the problem is when those senior folks retire.” There may be a talent shortage for experienced oversight.
Enterprises will undermine their AI efforts if there is little investment in training new talent to support AI’s future.
“We could actually be perpetuating [the very problem] we’re trying to solve with AI by not having that next pipeline of resources available to help with technology moving forward,” said Seewald.
