eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
According to the latest Appen survey published in October, artificial intelligence project deployment with meaningful return on investment (ROI) fell from a mean of 51.9 percent in 2023 to 47.3 percent in 2024.
Appen’s 12-page report, “The State of AI in 2024,” exposes the challenges enterprise AI faces moving forward, including data management, reliability of AI models, and scalability of AI development.
The study further highlights the factors influencing AI developments, one of which is the quality of data coupled with the complexities during the implementation that most companies adopting AI encounter.
Despite the 17 percent increase in the adoption of generative AI—56 percent in 2024, up from 39 percent in the previous year—ROI is down as a result of deploying low-percentage enterprise projects. This decline coincides with the same trajectory of enterprise AI projects making it to deployment, down from 50.9 percent last year to 47.4 percent this year.
It is not a new phenomenon that investment in generative AI and ROI are low in 2024. Last year, it was reported that investment in generative AI was a meager one percent of cloud spend, amounting to $2.5 billion. In the first half of 2024, Gartner’s May report revealed the same trend.
Appen’s study points out the main reasons for this downward trend: the importance of data quality, data management, and the expertise of data partners. Companies value diversity, bias reduction, and scalability, emphasizing that “human insight is key to refining AI systems,” 80 percent of the survey respondents said.
The bottleneck of data management, including data sourcing, cleaning, and labeling by humans, has increased by 10 percent. This is not to mention the lack of data availability, which also increased by seven percentage points, from 22 percent last year to 29 percent. Ninety-three percent of respondents say companies are looking for more efficient ways to manage data, and that requires data partners with expertise in the AI data lifecycle.
This decrease in enterprise deployment and ROI could only be temporary as companies are beginning to figure out ways to make the most out of the increasingly advancing AI technologies. Investment in enterprise AI applications requires time to blossom.