Making money isn’t everything…at least not when it comes to AI. A study by professional services firm Deloitte shows that for most companies, adopting AI tools has not improved their bottom line at all. But researchers continue to tout the merits of this technology.
According to Deloitte’s “State of AI in the Enterprise” report (PDF), 74% of organizations want their AI initiatives to increase revenue, but only 20% have seen it happen.
The firm’s conclusions echo a recent PwC survey of business leaders which reveals that only 12% of CEOs saw both lower costs and increased revenue from AI investments.
Deloitte’s explanation for the current situation is that money is not everything.
“AI success is not just about improving efficiency or even increasing revenue,” the report states. “It’s about achieving strategic differentiation and sustainable competitive advantage in the market.”
Business investments in AI have not been entirely in vain. Among the 3,235 global IT and business leaders who participated in Deloitte’s survey, 25% said AI was having a transformative effect on their organization, up from 12% a year ago.
When asked about the real benefits of AI today, 66% of respondents said it improves productivity and efficiency. How this works when only 20 percent report income growth remains unanswered. We note that a study published last year by METR, a nonprofit organization, found that AI coding tools made developers less productive, despite expectations to the contrary.
Even without a compelling financial reason, staff access to AI tools is growing. Less than 60% of workers now have access to IT-approved AI tools, compared to 40% a year ago. But among these AI-enabled workers, less than 60% use their AI tools as part of their daily workflow.
“This suggests that even as access is expanding, enterprise AI remains underutilized and its potential for productivity and innovation is still largely untapped,” the report speculates.
That said, it is likely that more AI pilots will enter production. Currently, 25% of organizations report having moved 40% or more of their AI experiences into real-world use. This figure is expected to reach 54% of organizations within three to six months.
The deployment of AI within companies appears likely to have an impact on employment, according to Deloitte.
“Within a year, more than a third of companies surveyed (36 percent) expect at least 10 percent of their jobs to be fully automated,” the report said. “The majority of companies surveyed (82 percent) expect at least 10 percent of their jobs to be fully automated within three years.”
This expectation, however, was not accompanied by major organizational changes. Around 84% of respondents said they had not redesigned roles based on AI’s capabilities.
People in these positions are also not convinced by AI technology. Among non-technical workers, only 13% are very enthusiastic about AI and are trying to use it, according to the report. While 55 percent are open to technology, 21 percent would prefer to avoid it and 4 percent are actively wary of it.
Ultimately, companies engaged in AI must convince their employees that the technology has benefits beyond automating tasks.
“Organizations that succeed with AI don’t just invest in automation and algorithms, they invest in their people,” Jim Rowan, U.S. AI leader at Deloitte, said in a declaration. “As AI continues to spark new ways of working, this dual focus – advancing both the capabilities of their talent and those of AI tools – empowers teams to adopt reimagined business models and lays the foundation for competitive advantage.
Another concern is “sovereign AI,” meaning that companies control their AI software and data in accordance with local laws and regulations and do not rely on foreign suppliers or infrastructure. Eighty-three percent of companies surveyed say sovereign AI is at least moderately important to them, and 43% say it is very or extremely important.
As for agents – AI models with access to tools – their use is currently modest, but is expected to increase. Today, 23% of companies say they use agents at least moderately, and within two years, that figure is expected to rise to 74%.
This slow adoption could prove beneficial, as only 21% of companies report having a mature governance model in place for autonomous agents.
Ali Sarrafi, CEO and co-founder of Kovant, an enterprise agent platform, said The register in an interview, the problem with the way people use AI is that they see it as a form of sophisticated workflow automation.
“There are studies that show that personal productivity doesn’t go that far if you do this,” he said. “People start using it. But as soon as they get tired of it, they go back to the way they did things before.”
The big shift, he said, when companies start seeing revenue results, comes from giving AI agents work as if they were colleagues and having them run automatically.
“We work with this big manufacturing company,” Sarrafi explained. “They have about 7,000 suppliers. And every time they needed to restock something, they had to coordinate with a lot of suppliers. It’s actually the most boring job for everyone. But then they deploy this agent or a team of agents who basically monitors the stock levels. this and at what price?
The result is a summary report sent to Microsoft Teams for a business planner to review and approve. If authorized to do so, Sarrafi explained, the agent then sends the purchase order and follows up with the supplier until the goods reach the warehouse.
“So all that manual, boring work, all of a sudden, saves about 95 percent of that,” he said.
Turning to Deloitte’s report, he said the consultancy’s focus on governance can be factored into product design – carefully designing AI workflows.
“They make governance a big deal, but in reality you have to have an agile governance model,” he said. “It’s the same way that when you hire people, you create governance around them. The classification of information needs to be addressed from the start. If you just open up the whole world to the agent, then of course it’s a statistical model – that could create problems. So it’s more of a design issue in my mind than a need for a massive governance architecture.”
Sarrafi also said Deloitte’s findings on worker hesitancy to AI can be attributed in part to cumbersome enterprise tools. “Most of the enterprise AI tools that are built, applications that are built for employees, actually don’t live up to what they expect in terms of user experience,” he said, adding that people don’t want to switch between multiple tools.
Successful implementations, he said, tend to allow users to interact with agents through familiar tools such as Microsoft Teams or Slack.
“I’m not going to name any products,” Sarrafi said, “but most enterprise AI tools right now are about a year or two behind consumer tools in terms of user experience.” ®