Survey: AI adoption continues apace, despite uncertainties and failure risk

Around 94% of manufacturing organizations are using some form of AI as adoption expands beyond automation and into predictive, generative, and connected applications despite supply chain uncertainties and implementation challenges or outright failures, according to a new survey from a Silicon Valley software and digital transformation stakeholder. 

See also: Crystal Ball 2026 Series

Despite uncertainties in supply chains and turmoil wrought by the Trump tariffs, manufacturers are continuing to prioritize AI adoption in digital transformation, according to San Ramon, California-based cloud ERP vendor Rootstock Software’s 2026 State of Manufacturing Survey.

Rootstock's survey polled 520 professionals who lead digital transformation at midsize to large manufacturers across a wide swath of geography, North America, Europe and Asia, and the results from it align with others from surveys by companies in the manufacturing technology and software space such as Loftware and Smart Industry's own recent State of Initiative study for 2026.

“The [94%] statistic illustrates that AI has a place in manufacturing, as it's now nearing a universal level of adoption,” said Ohad Idan, Rootstock's VP of product. “It’s about investing in understanding what works for businesses so that they can reliably get meaningful results.” 

AI is blossoming but in still-to-be-determined directions for manufacturing

The results follow a trend of manufacturers and other companies pouring record investments into new technology. Private investment in the U.S. in AI reached $109.1 billion in 2024, compared to China’s $9.3 billion and the U.K.’s $4.5 billion. 

Rootstock’s findings echoed this level of commitment, as the company's survey showed that 61% of manufacturers plan to increase enterprise software spending over the next 12 months, with the largest share planning increases of 11% to 25%. 

That trend suggests measured growth that’s focused on modernization and efficiency gains, according to the survey. 

Rootstock found that amid tariffs and global trade pressures, companies are investing in forecasting, with 39% expecting higher raw material costs and 37% planning to raise prices, increasing focus on tariff management, sourcing strategies, and cost forecasting. 

For example, the survey found that predictive AI saw the largest adoption gain at 48%, while AI investment shifted most sharply to supply chain planning at 35% and process optimization at 36%, signaling rapid growth in forecasting and content-driven use cases.  

See also: How data, governance and organizational change define AI success 

This is similar to findings in Smart Industry’s State of Initiative Survey, which found almost 75% of respondents saw drastic challenges with their incoming supply chains (the kind most effected by tariffs) while 54% responded that they saw at least "some" challenges with their incoming supply chains.

However, AI implementation does not come without challenges.

Despite the spending and adoption, most AI remains in experimentation or piloting stages, and failure rates are high. According to reporting by Smart Industry, 70% to 85% of Gen-AI deployments fail to deliver their intended ROI, which is double the already-high failure rate of traditional IT projects.

See also: Executives advocate for reshaping of workforce following job cuts from AI  

At the same time, regulatory pressure is rising. In 2024, 59 AI-related regulations were introduced by U.S. federal agencies, which was double the prior year.

Meanwhile, supply chain unreliability persists and tariffs continue to create uncertainty. For example, Smart Industry’s survey found that last year, 69% of respondents reported drastic challenges while an almost identical number to this year, 55.79%, reported some challenges. Interestingly, this year 62% reported no changes or challenges to their incoming supply chains, while 51% said the same thing last year. 

At the same time, almost 26% saw drastic challenges with their outgoing supply chains (the number last year was 31%). 

Rootstock also found that companies frequently faced challenges with digital transformation projects constrained by internal organizational factors, and Smart Industry’s own results yielded the same sentiment.

The Rootstock survey found that 33% cited lack of the right talent as a primary barrier, while 31% report insufficient cross-department collaboration, highlighting that people and operating-model challenges are key inhibitors to progress. 

But Idan said that despite these challenges, AI adoption is still worth it for companies even with risks of failure.  

While the Rootstock survey did not specifically ask manufacturers about failures, it did measure perceived AI maturity, which is what companies think about the success of their AI projects, finding that 73% of manufacturers now believe they are "on par" with or "ahead" of peers in AI adoption.

This is important, Idan said, because “their perception is they’re not falling behind their peers,” so despite risks of failure, it's still worth it for companies to adopt the technology. 

See also: Predictability at scale: How AI and automation can transform IT operations

He added that there is still a gap in AI leadership, and it is unclear how AI can differentiate organizations. Additionally, failure with one program doesn’t translate to failure overall, as companies continue to use AI across different programs and platforms to determine how it can fit into their needs.  

Idan emphasized the importance of a “unified platform” as manufacturers look to ERP as a way to consolidate platforms and bring fragmented parts of the business together. 

“Manufacturers are increasingly taking a unified platform approach as it reduces the need for complex integrations, consolidates data, and provides greater visibility,” he said. “In this type of environment, AI is able to access data from the entire business lifecycle and can take actions across domains.” 

See also: Poka leadership transition signals shift toward AI in its connected worker platform 

He also highlighted some ineffective approaches that businesses have, such as when companies rely on technology and adopting it without consideration of business problems, as confirmed by Smart Industry’s Crystal Ball Series. 

He said the approach needs to start with desired business outcomes then determining the best technology to be implemented, rather than relying solely on technology to solve the problem.  

“There is not one clear path for AI adoption to make manufacturers more successful,” he said.