- Enterprise leaders devote 58% of their AI budget to the data sourcing and preparation stages, according to Appen's 2022 State of AI report, published Tuesday. The company surveyed 504 IT leaders in the U.S. and Europe.
- The rest of the AI budget goes toward model testing and deployment, and human evaluations of modes, according to the report.
- Research indicates there's a difference in how business leaders and technologists perceive the earlier stages of AI development. While 42% of technologists find data sourcing very challenging, just 24% of business leaders share the sentiment, the survey found.
AI's success hinges on access to clean, valuable data which can train models accurately to make decisions or speed up operations. This is often the hardest part of the process, especially as companies contend with a multitude of potential data sources.
Nearly one-third of survey respondents point to a dearth of data as a key bottleneck of AI objectives. The top issue, cited by more than two in five respondents, was data management.
But business leaders and technologists struggle most with AI at different parts of the process. The mismatch can lead to a "misalignment in priorities and budget within the organization," according to the report.
A Forrester report shows businesses with higher proficiency in using insights gleaned from data are increasingly drawing data from external sources, particularly data brokers.
But even as AI technology becomes easier to infuse into the organization, there are challenges to successful adoption. A new Gartner report shows AI is infused into a wider number of enterprise products and services, which in turn can lead to fragmented AI strategies. Additionally, companies are struggling to find, attract and retain the talent needed to carry out AI programs.