Many firms have no formal data strategy despite increasing AI usage, according to a new study from Carruthers and Jackson.
More than a quarter (26%) lack a data strategy, the study found, while 39% reported little to no data governance throughout their business.
These same firms are still pushing on AI and ramping up their adoption efforts, though. Just 7% of businesses now go without AI, falling from 26% of businesses a year ago, while 53% reported an increase in the use of AI tools.
There’s also a lack of data literacy among staff, with 57% reporting a deficiency in this understanding among employees.
Similarly, while 44% have seen an increase in conversations around ethical AI, only 13% have turned these conversations into structured policies around AI ethics.
The report revealed some positives, however, with 37% claiming their business adopts multiple governance frameworks around AI. This figure is up from 31% in 2023.
“An AI Paradox has been created, as the use of AI tools in organizations has surged in the last year, yet employees lack the data literacy to use them effectively, as their fundamental understanding of data remains largely unchanged from last year,” Caroline Carruthers, CEO of Carruthers and Jackson, said.
Richard Bovey, data chief at AND Digital, said that there is a worrying knowledge gap around the risks of rushed implementation and that businesses need solid data governance structures to cover technology development.
“That’s especially true when it comes to AI, not only facilitating greater accuracy and reliability within AI outputs but helping to mitigate potential privacy and security concerns,” Bovey said.
“Weaving data governance principles across the development and integration process, supported by data skills, will greatly increase the likelihood of success for AI systems, providing a platform for them to be a business enabler rather than a costly mistake,” he added.
Good AI needs good data
It’s becoming increasingly clear to firms that data is of central importance to AI adoption projects, with studies pointing to the importance of data quality.
In a recent report from Ataccama, 68% of chief data officers claimed that data quality was their top challenge, while 40% said they struggle to maintain consistent data quality. This directly hinders outcomes, the report found.
“At the heart of every successful AI deployment is a foundation of high-quality, well-managed, and readily accessible data, so it’s concerning to see businesses ramping up AI usage without the basics in place,” Stuart Harvey, CEO of Datactics, told ITPro.
“The outputs of AI models are only as good as the data they are trained on and input, so disregarding a robust data strategy is only setting up for failure,” Harvey added.
Data storage may start to cause problems, as well. Research from Hitachi Vantara at the start of last year found that data storage capacity requirements were set to surge by 150% over the next few years.
