According to research from the Enterprise Strategy Group (ESG), IT leaders are increasingly prioritizing generative artificial intelligence (GenAI) initiatives aimed at boosting productivity and operational efficiency as the primary advantages of early technology adoption.
Despite a significant increase in investment, with 70% of organizations boosting their GenAI spending over the past year, fewer than 10% of IT decision-makers surveyed report having implemented GenAI applications in a live environment. In ESG’s global survey of 832 IT leaders for its “State of the Generative AI Market” report, only 8% of respondents classify their GenAI projects as being in “mature production.” However, this number has doubled over the past year, and the percentage of organizations in early production has increased by over a third (36%). Consequently, 27% of the executives surveyed work in companies where AI is either in early or mature production. Moreover, the number of organizations conducting pilots and proofs of concept has seen a 22% increase compared to 2023.
ESG is a subsidiary of TechTarget, which publishes Computer Weekly, and the survey included IT leaders from Computer Weekly’s global audience.
### The Impact of GenAI on IT
Currently, the IT department benefits the most from GenAI implementations. On average, IT decision-makers are utilizing GenAI across 3.5 application areas, with software development emerging as the top application—41% of respondents indicate its usage. This marks a 7% rise from the previous year.
PwC has identified several use cases for GenAI in software development, including automated test script generation, detailed troubleshooting, code reviews, code completion, and automated documentation. They predict that skilled users will soon be able to direct GenAI to produce high-quality artifacts for user stories, acceptance criteria, test scenarios, documentation, and even automatically generate APIs, ultimately enhancing every facet of the agile software development lifecycle.
Other areas witnessing notable increases in GenAI applications since 2023 include research, IT operations (up 7%), and cyber security (up 6%). The broader application of AI to automate manual-intensive processes within IT—termed AIops—is regarded by many as a strategy for navigating increasingly complex IT environments. For instance, energy provider EDF is employing Dynatrace’s AI-based automation technology to optimize its cloud operations and ensure consistent, secure customer experiences. They view Dynatrace’s AI monitoring as a solution for spotting inefficiencies in their tech infrastructure, enabling swift downtime remediation and improved customer self-service.
Regarding cyber security, the Alan Turing Institute’s Centre for Emerging Technologies has discussed how GenAI introduces both new risks and opportunities. According to principal research engineer Sarah Mercer and director Tim Watson, GenAI could modify the cyber security landscape in various ways. While it may heighten existing risks related to reconnaissance, social engineering, and spear phishing, its code-generation capabilities appear to have a more muted effect on the threat spectrum. Mercer and Watson emphasize that existing GenAI systems excel particularly in pattern recognition and natural language processing, harnessing extensive training data and offering multimodal capabilities. “Focused application of these strengths could significantly enhance current technologies for both cyber offense and defense,” they noted.
### Maturity in Enterprise GenAI
Organizations with mature AI deployments are progressing more rapidly, particularly in the use of GenAI for software development (10%), IT operations (15%), and finance (11%) compared to the global averages. These mature enterprises are also more inclined to leverage generative AI for managing cloud infrastructure expenses.
One of the benefits of GenAI is that it can be trained on publicly available internet data, which allows for quick deployment of valuable AI applications. However, this universality means that organizations have access to similar large language models (LLMs) trained on the same data, diminishing any competitive edge that might be achieved through GenAI.
The ESG survey indicates that IT decision-makers are aware of the limitations of public GenAI models and understand the importance of supplementing these models with their own data or creating internal LLMs tailored to their specific business needs. Although extensive public datasets have facilitated GenAI’s rise, 84% of IT leaders believe it is crucial for their organizations to utilize proprietary data to support GenAI initiatives. Furthermore, 77% emphasize the importance of training their own GenAI models. This is reflected in the number of organizations that deploy multiple LLMs for generative AI, with two-thirds of IT leaders using two or more LLMs, and nearly 10% employing five or more.
