Manufacturing leads peers in AI-driven IT issue fixing

TeamViewer has published research on AI adoption among manufacturing IT decision-makers, finding the sector ahead of its peers in using proactive and partially autonomous approaches to resolving technology issues.

Among manufacturing respondents, 43% said their organisation already resolves issues proactively, while a further 25% reported partially autonomous handling, where AI predicts and prevents issues with some human oversight. That gives the sector a combined total of 68%, ahead of technology at 56%. Across all industries, the figures were 41% for proactive handling and 17% for partially autonomous handling.

The findings suggest manufacturing is relatively open to using automation in day-to-day IT operations, but less certain about how to measure its broader impact. Almost a third of manufacturing IT decision-makers, 32%, said uncertainty about AI's value or business impact remained a barrier to adoption, compared with 16% in technology and 21% across all industries.

Security, privacy and risk concerns were the most commonly cited obstacle in manufacturing, identified by 47% of respondents. A lack of skills or training followed at 34%, placing those issues alongside doubts over business value as the three main barriers to wider uptake.

Proactive preference

The survey also examined how manufacturing staff want IT support to operate. In the sector, 28% said they would prefer IT teams to use AI to detect issues early and keep them informed, slightly above the 25% cross-industry average.

By contrast, only 19% said they would rather report issues themselves and wait for a fix. That was notably lower than the 25% cross-industry average, suggesting a stronger preference in manufacturing for intervention before problems become visible to end users.

Views on human involvement also differed from the broader sample. Just 23% of manufacturing respondents said a person should be involved before any action is taken, compared with 31% across industries. At the same time, 46% said they wanted a human involved when security, data or business operations could be affected, above the 40% average.

This suggests greater acceptance of AI for monitoring and routine intervention, while preserving human oversight in higher-risk situations. Across the wider research, respondents were more comfortable with visible and assistive tasks such as system monitoring, issue flagging and transcription than with AI making independent decisions or handling sensitive data without supervision.

Productivity gains

Manufacturing respondents also reported a stronger personal productivity effect from AI than the market average. Some 44% said AI is already making them more productive, compared with 39% across industries.

They also spent less time checking AI outputs than peers in other sectors. Manufacturing respondents who verify AI outputs said they spend an average of 1.6 hours a week doing so, compared with 1.9 hours across industries and 2.7 hours in technology.

Daily use, however, was somewhat lower than in other sectors. Manufacturing respondents said they prompt AI an average of 3.08 times a day, compared with 3.28 across industries and 4.72 in technology.

That combination of lower prompting frequency and higher reported productivity suggests AI may be used more selectively in manufacturing environments, particularly for issue detection and operational support. The study did not indicate that this had resolved questions over return on investment at organisational level, where doubt remains more pronounced than in technology companies.

Looking ahead, 35% of manufacturing respondents said they expect to delegate more tasks to AI by 2030 and focus on higher-value work. That compares with 28% across industries, pointing to continued appetite for a broader operational role for AI within manufacturing IT functions.

The research also found that comfort levels with autonomous AI varied by working pattern. The gap was widest among onsite and hybrid employees, who were less comfortable with autonomous AI than remote employees.

Mei Dent, Chief Product and Technology Officer at TeamViewer, said the challenge in manufacturing often lies in proving the value of changes that happen incrementally rather than through full system replacement.

"When people talk about AI in manufacturing, they often imagine fully autonomous environments. In reality, many manufacturers run on systems built over decades that they can't simply replace, so the biggest advances come from thousands of smaller decisions: identifying issues earlier and removing everyday friction. Because that work happens quietly in the background, its impact is easy to overlook. The organisations creating the most value are often those solving these small problems at scale and consistently demonstrating the results in specific use cases," Dent said.

The study was conducted by Sapio Research on behalf of TeamViewer among 4,200 respondents across nine markets: the US, the UK, Australia, France, Germany, Italy, Japan, Singapore and Brazil.