SoftServe study shows global gap in generative AI success

SoftServe has released findings from a recent study evaluating the use of Generative AI (Gen AI) across global businesses, including those in Singapore.

The study was conducted by Forrester Consulting and surveyed 777 technology purchasing decision-makers involved with Gen AI in their organisations.

The study revealed that while enthusiasm for Gen AI remains high, only 22% of organisations report effectively using the technology across all business functions. More than half of global decision-makers have established business goals for using Gen AI, yet 79% are concerned about their organisation's ability to execute these goals with their current levels of expertise. This concern is echoed in Singapore, where similar levels of success and challenges were reported in unlocking Gen AI's value.

"Despite a swift start to the Gen AI race, many initiatives get stuck in the piloting stages as more organisations realise their data infrastructure isn't ready to adequately deploy Gen AI technologies beyond the proof-of-concept," said Alex Chubay, SoftServe's Chief Technology Officer. "Gaps in skills and knowledge of emerging Gen AI technologies, technical feasibility, and data readiness hinder companies from moving beyond tactical wins in pilot mode to full-scale deployments enabling novel business capabilities and experiences."

The study found that 42% of global organisations have the capabilities to train Gen AI models, while 89% face difficulties in preparing business data for Gen AI use. In Singapore, only 35% of firms manage to train their models on enterprise data, and 51% embed the data into their models instead. Furthermore, less than one-fourth (24%) of organisations have governance plans in place, even though 90% agree that adopting a governance plan is crucial for responsible use and risk mitigation of Gen AI.

The gap between expectations and reality is apparent as only 3% of organisations' models can leverage a full range of six or more types of business data, such as operational, customer, employee, source code, public, and partner data. The average number of data types used is just three. Additionally, 88% of respondents noted the increasing importance of technical expertise for data integration, model optimisation, use case development, and further application development.

According to the study, companies are seeking external expertise to accelerate deployment and gain a better understanding of industry-specific needs. A significant 80% of decision-makers reported that their employees struggle with use case awareness and understanding of Gen AI's complexity, while 90% indicated a need for partners with advanced technical capabilities to realise transformative value in future use cases. Companies are looking for partners who can provide accelerated deployment support (89%) and industry-specific understanding (88%) to aid in execution and implementation.

The study identified notable trends in Gen AI results, including significant value for organisations prioritising data, governance, and skill development with support from technical partners and experts. The U.S. led in unlocking Gen AI value, followed by the UK, Singapore, and Germany. Retail was most proficient in harnessing Gen AI value and training models on owned data, while the financial services and insurance sectors encountered more challenges before yielding any gains. These sectors also reported releasing fewer governance plans than others.

Healthcare, life sciences, oil and gas, manufacturing, ISVs, and enterprise technology sectors showed an even divide in achieving Gen AI value. Larger businesses with revenues greater than USD $5 billion were less likely to demonstrate Gen AI success due to difficulties in organising the needed capabilities across expansive hardware, software, and infrastructure landscapes.