BMC Software survey: IT leaders bullish on generative and agentic AI

Generative and agentic artificial intelligence (AI) have captured the imagination of IT leaders, but there is a significant gap between enthusiasm and implementation maturity for IT operations and service management, according to a new survey from BMC Software and Dimensional Research. The study polled 424 IT stakeholders responsible for some aspect of AI implementation across large and midsize organizations in a wide array of industries.

Even though 94% said that generative AI (genAI) is a part of their operations and management strategy, only 17% of the participating organizations have moved beyond the experimental phase, with just 5% reporting that they have mature implementations. The majority remain in the initial stages: 41% are experimenting with simple use cases, and 42% are running initial pilots.

The adoption pattern reveals complexity in implementation approaches. Rather than settling for a single solution, organizations are embracing multiple platforms:

• 44% use two different genAI solutions, and 22% employ three.
• 71% use Microsoft Copilot, 62% use OpenAI ChatGPT, and 32% use Google Gemini.

Given the complexity of training and managing a large language model (LLM), 24% are utilizing open source LLMs such as Llama or Google Gemini and developing their own homegrown solutions.

Use cases for genAI are varied, although chatbots lead the way (see chart).

IT faces many significant challenges in implementing AI, according to the survey. Nearly half of the respondents (49%) said data is scattered across multiple repositories, and 45% reported that their existing solutions focus on simple, reactive use cases that don’t tap into AI’s full capabilities.

There’s also a knowledge gap: 43% said that business stakeholders don’t understand the distinction between generic generative pretrained transformers (GPT) models and specialized genAI solutions for IT operations.

Yet, optimism remains high, with 98% of the respondents expecting to reap future benefits from genAI, such as increased productivity, reduced maintenance needs, and lower costs. Also, the more mature the organization’s AI implementations are, the more optimistic the respondents are about the technology; 62% of those with advanced implementations predict significant future value.

The survey also explored attitudes toward emerging agentic AI capabilities, which revealed another discrepancy. Although 60% of the respondents cited the value of natural-language interfaces for analytical reporting and 58% acknowledged the potential of autonomous agents, familiarity with agentic AI remains low. Only 8% consider themselves very knowledgeable about the technology, and 37% have never heard of it.

Finally, contrary to reported concerns that AI will replace workers, 63% of the responding service and operations stakeholders believe that genAI will improve jobs, with only 1% expressing concerns about job displacement. This suggests that IT professionals see AI as an enhancing force rather than as a replacement for human expertise.

IT is at a critical inflection point when it comes to generative and agentic AI. The survey results highlight the need for organizations to both embrace AI’s potential and recognize the practical challenges for implementation. Once these challenges are overcome, however, the next phase of AI adoption will likely focus on more sophisticated applications that continue to transform IT operations.

Download the full report or attend this webinar to learn how agentic AI will transform enterprise IT work for your organization, and contact BMC today.