AI scaling surges but enterprises struggle with persistent and emerging challenges

In the second instalment of this blog series, Omdia Chief Analyst Eden Zoller examines the state of AI maturity across enterprises and explores real-world use cases, key barriers to adoption, and the factors influencing organizations’ readiness to fully leverage AI's potential. 

Omdia’s 2024 AI Market Maturity Survey* reveals that nearly a quarter of enterprises have reached an advanced stage of AI maturity, scaling AI across multiple use cases at higher levels than in 2023. The current survey reveals striking and often surprising insights that will prompt serious reflection among enterprises and vendors alike. High priority AI investment drivers such as cost savings score low when it comes to exceeding expectations, while many low priority investments are smashing expectations. Enterprises are finding security and data privacy more, rather than less problematic in 2024 while concerns over AI compliance are intensifying. Moreover, comprehensive senior management support for AI is fragmented, with almost half of enterprises reporting  uneven levels of buy-in. 

Large enterprises driving AI at scale: 37% of firms with revenues above $1 billion have scaled AI across multiple business functions and/or units, in contrast to 13% for enterprises with revenues below  $250 million.  This is logical as large firms tend to have better data infrastructure, in-house expertise, and bigger budgets in invest in AI technology. When it comes to scaling AI across specific use cases, in 2023 13% was the highest level reported, for customer experience use cases. In 2024, 23% is the highest overall level of scaling, associated with sales and marketing use cases. Several other use cases are scaling at similar levels, such as customer support (21%), product/service personalization and cybersecurity. 

Generative AI (GenAI) is driving AI investments for just over a third of enterprises in the survey, with as many again saying GenAI is exceeding expectations. However, other priority AI investment drivers  not delivering such stellar levels of satisfaction. Where cost savings are a priority investment driver for 33% of respondents, only 14% say that AI deployments in this area are exceeding expectations.  Efficiency gains are an AI investment driver for 33% of respondents, but only 18% report such investments have exceeded expectations. The findings are similar for the use of AI to enhance customer experience management, which taken together are sobering messages for vendors with solutions in these domains. In contrast many of the lower priority investment drivers for AI are exceeding expectations to a high degree. For example, the use of AI to assist R&D and speed up product development is only an investment priority for 11% of enterprise, and yet 46% of enterprises say that AI in this capacity is exceeding expectations. Harnessing  AI to create new products and services is a priority for 20% of enterprises in the survey but results here have exceeded expectations for a third of respondents. The latter results are impressive, particularly as benefits from using AI for R&D, and to create new products new takes time to manifest, unlike cost savings and efficiency gains that can have more immediate impacts. 

Barriers to AI adoption have intensified for enterprises in the 2024 survey. AI compliance and regulatory challenges are proving problematic for 35% of respondents compared to 11% in 2023. The increase reflects the introduction of new laws for AI, with more on the horizon. Security and data privacy issues are the biggest concern among 2024 survey respondents (43%). Security and data privacy and are an ongoing issue for AI that have been intensified by aspects of GenAI. The large datasets for GenAI models are typically obtained from web scraping, which can absorb personal data that is not always protected by strict privacy controls. In addition, models can be vulnerable to data leakage and memorization, which impacts data privacy. Lack of qualified staff and in house expertise is an issue for a third of respondents in the 2024 survey, compared to 18% in 2023.  AI adoption requires a unique blend of technical skills that can be hard to obtain (e.g., data scientist, programmers and engineers with machine learning expertise). The problem is most acute for the for smaller firms in the survey with revenues below $250 million (39%).  Poor data availability and quality is a widely recognized challenge for AI development and is an issue for over a quarter of enterprises in the survey. AI models rely on high-quality, relevant, and well-structured data. But many enterprises still have data that is fragmented across departments and functions, stored in inconsistent formats and/or prone to inaccuracies and missing values. AI deployments will be inadequate without a solid data foundation, which includes data governance and privacy, centralized infrastructure for data storage, cleansing and processing. The complexity of AI and integration problems are a significant challenge for survey respondents, with notable spikes in the energy and utilities vertical (48%), telecoms (42%), media and entertainment (38%) and regionally in Africa (50%). Pilot projects can be used to explore and navigate AI integration issues ahead of wider deployments. Middleware solutions can be useful, acting as a layer that bridges old and new systems, and orchestrates workflows across systems. Modular, flexible AI solutions that are data agnostic (i.e., can handle multiple data sources and formats) and that come with well-documented APIs can also help address integration issues. 

Although 44% of survey respondents say that AI is championed at the highest management levels, 49% report more fragmented management support for AI – by some managers and/or departments but not all of them. Senior management support is notably lower for firms piloting AI (36%) and for those exploring AI technology (31%). Senior management should support AI in a sustained way through all phases of AI deployment, not just at maturity. The early phases of AI deployment are when an enterprise need more, not less support. This is the phase where senior executives should be leading efforts to define AI strategy, goals, projects and ROI. Strong senior management buy-in from the start is also critical in securing decent AI budgets that enterprises need to get AI projects off the ground and for the overall success of AI initiatives. 

* The Omdia 2024 AI Market Maturity Survey was completed in August 2024 among 478 enterprises from across the major global regions, key industry verticals, and companies of different sizes based on revenue.