New research commissioned by ABBYY reveals the major challenges faced by manufacturing businesses trying to integrate sophisticated Generative AI (GenAI) technology.
The survey, conducted by Opinium, revealed that 34% of manufacturing leaders said staff struggled with deploying it due to a lack of skill, 31% found training the models harder than expected and 30% had difficulty integrating with current business processes.
About 21% lacked an AI policy or governance. Managers also complained that staff misuse the tools (23%).
In response, many business leaders addressed these challenges by combining other technologies, with 40% turning to AI agents to improve outputs, 34% turning to process intelligence and 25% to RAG, according to the 2025 ABBYY State of Intelligent Automation Report: GenAI Confessions. Nearly 53% also invested in staff training.
Adding these technologies contributed to 98% of manufacturing respondents ultimately being happy with their GenAI tools. As a result, 57% saw output quality and consistency improve and 41% saw cost savings. Some 43% saw more accurate and reliable results and 44% saw better integration into existing systems.
Despite this optimism, however, the average budget for GenAI is only expected to increase by 18% in 2026, compared with 24% in financial services and 22% in Transport & Logistics.
Evaluate processes before investing in GenAI
"Businesses are spending money on GenAI tools that promise more than they can provide," Senior Director of AI Strategy at ABBYY Maxime Vermeir said. "In some cases, they don’t even need it.
"Before moving forward with leveraging GenAI tools or agentic AI, companies need to first evaluate their current processes and create a visibility map of their workflow with data analytics tools such as process intelligence. When training models prove more difficult than expected, pre-trained, purpose-built AI turns out to be the right solution."
Staff potentially putting businesses at risk with unauthorized use of GenAI
The survey also revealed that more than a third (45%) of manufacturing leaders admit that a driving factor for introducing GenAI was that employees were already using it on a Bring Your Own Software (BYOS) basis for personal productivity, which could impact security concerns over ‘Shadow AI’.
"When individuals use commonly available tools like ChatGPT, Grok or Perplexity without oversight at work, [it] potentially raises serious data privacy and compliance concerns," ABBYY CEO Ulf Persson said. "The corporate benefits of GenAI’s potential is truly unlocked when leaders drive secure, strategic adoption with risk management as a priority.”
Other key findings
Goals for GenAI use include increasing efficiency and customer service (64%) and improving on existing results (55%)
58% use GenAI for data analysis and insights generation, 51% for customer service and the same proportion for employee productivity, 49% for sales and marketing optimization and 48% for automating document business processes (such as accounts payable)
When considering improvements to GenAI, top of manufacturing firms’ wish-list was that GenAI would save time through manually processing information from documents (32%). This was followed by knowing which processes could be improved and improve them (30%) and GenAI asking additional questions to provide a more accurate result (26%).
