59% of organizations lack resources to meet generative AI expectations: study

A recent study conducted by an open-source AI solutions company ClearML in collaboration with the AI Infrastructure Alliance (AIIA) has shed light on the adoption of generative AI among Fortune 1000 (F-1000) companies.

The study, “Enterprise Generative AI Adoption: C-Level Key Considerations, Challenges, and Strategies for Unleashing AI at Scale,” revealed the economic impact and significant challenges C-level executives face in harnessing the potential of AI within their organizations.

>>Don’t miss our special issue: The Future of the Data Center: Coping with Growing Demands.<

According to the global survey, 59% of C-suite executives lack the necessary resources to meet the expectations of generative AI innovation set by business leadership. Budget constraints and resource constraints emerged as critical barriers to successful AI adoption within enterprises, hindering the creation of tangible value.

The survey also found that 66% of respondents cannot fully measure the impact and return on investment (ROI) of their AI/ML projects at the bottom of the line. This highlights the profound inability of underfunded, understaffed, and undergoverned AI, ML, and engineering teams in large enterprises to effectively quantify results.

“While most respondents said they need to scale AI, they also said they lack the budget, resources, talent, time and technology to do so,” Moses Guttman, co-founder and CEO of ClearML, told VentureBeat . “Given AI’s multiplier effect on revenue, new product ideas and functional optimization, we believe companies now need to allocate critical resources to invest in AI to effectively transform their organization.”

The study also highlights rising revenue expectations from AI and ML investments. More than half of respondents (57%) report their boards expect double-digit revenue growth from these investments over the next fiscal year, while 37% expect single-digit growth.

The study collected responses from 1,000 C-level executives, including CDOs, CIOs, CDAOs, VPs of AI and digital transformation, and CTOs. According to ClearML, these executives are leading generative AI transformation in Fortune 1000 and large enterprises.

The State of Generative AI Adoption

According to the survey, most respondents believe that unleashing AI and machine learning use cases is critical to creating business value. Eighty-one percent of respondents named it a top priority or one of their top three priorities.

Additionally, 78% of enterprises plan to adopt xGPT/LLMs/generative AI as part of their AI transformation initiatives in fiscal year 2023, while another 9% plan to begin adoption in 2024, bringing the total to 87%.

Respondents were also nearly unanimous (88%) on their organization’s plan to implement policies specific to the adoption and use of generative AI across all business units of the enterprise.

While generative AI and ML adoption is a key driver of enterprise revenue and ingenuity, 59% of C-Level leaders lack the resources to meet business leadership expectations for gen AI innovation.

They face budget and resource constraints that hinder adoption and value creation. In particular, people, process, and technology are all critical pain points identified by F-1000 and enterprise executives when it comes to building, running, and managing AI and machine learning processes:

• 42% say there is a critical need for talent, especially skilled AI and machine learning staff, to achieve success.
• Another 28% flag technology as the main barrier, indicating a lack of a unified software platform to manage all aspects of their organization’s AI/ML processes.
• 22% cite time as a major challenge, describing the excessive time spent on data collection, preparation and manual pipeline construction.

In addition, 88% of respondents said their organization is aiming to standardize on a single AI/ML platform across departments rather than using different point solutions for different teams.

“Enterprise decision makers are poised to invest more in generative AI and ML this year, but according to our research results, they are looking for a centralized end-to-end platform, not spreading spending across multiple point solutions,” ClearML’s Guttmann told VentureBeat. “With a growing interest in realizing business value from AI and ML investments, we expect demand for greater visibility, seamless integration and low-code to drive generative adoption of AI.”

Key challenges hindering generative AI adoption

The research found that growing concerns about AI and generative AI governance have led to serious financial and economic consequences.

It found that 54% percent of CDOs, CEOs, CIOs, heads of AI, and CTOs reported that their failure to manage AI/ML applications resulted in losses to the enterprise, while 63% of respondents reported a loss of $50 million or more due to inadequate management of their AI/ML applications.

When asked about the top challenges and barriers to adopting generative AI/LLMs/xGPT solutions within their organization and business units, respondents identified five key challenges:

• 64% of respondents expressed concerns about customization and flexibility, particularly the ability to build custom models using their new internal data.
• 63% of respondents cited data retention as a top priority, with a focus on generating AI models and protecting business intelligence to maintain a competitive advantage while protecting corporate IP.
• 60% of respondents highlighted governance as a major challenge, stressing the importance of limiting access to and managing sensitive data within the organization.
• 56% of respondents said security and compliance were top-of-mind, given that companies rely on public APIs to access generative AI models and xGPT solutions, exposing them to potential data breaches and privacy concerns.
• 53% of respondents cited performance and cost as one of the top challenges, primarily related to fixed GPT performance and associated costs.

According to Guttmann, the lack of visibility, measurability and predictability identified in the research poses a difficult obstacle to success in adopting new technology. All of these factors are critical to success.

“Enterprise customers should aim for out-of-the-box LLM performance, trained on their internal business data securely on their on-prem installations, resulting in cloud cost reduction and better ROI,” he said.

At VB Transform, ClearML unveiled a new Enterprise Cost Management Center. This center enables enterprise customers to efficiently manage, predict and reduce rising cloud costs.

In addition, the company plans to release a calculator to help companies understand and predict their total cost of ownership and the hidden operating costs of gen AI. ClearML said this tool will provide valuable insights for better cost management and informed decision making.

.