Fears that poor data could scupper the effective adoption of artificial intelligence systems appear to have been confirmed, with most businesses struggling to manage or access quality data, despite pouring more and more resources into AI tools.
So says a new study from SoftServe, which exposes a growing gap between companies’ AI ambitions and their data capabilities.
The rapid deployment of AI tech has made data a critical resource for companies looking for an edge on their competition, however, the study has found a huge gap between business leaders’ goals and their ability to use data to deliver technological transformation.
“The Great Data Divide” report has found that while almost all (98%) business and tech leaders believe that a data strategy update is required to gain the full benefits of AI, 73% admitted their company has allocated funds or talent to GenAI trends at the expense of more valuable data and analytics initiatives.
In fact, nearly three-fifths (58%) said that their company had made strategic business decisions based on inaccurate or inconsistent data most of the time, while even more (60%) said it was a real challenge to get access to the data they need, when they need it.
SoftServe’s report, based on a survey of 750 business leaders, as well as technology and data executives, found that a two-thirds (65%) of firms said no one in their organisation has a comprehensive understanding of the data collected or how to access it.
An overwhelming majority of business leaders (98%) confessed that they were facing “considerable challenges” in improving their firm’s data strategies, with the main barriers being the use of legacy or varied tech by different units (73%) and data ownership not being consolidated (75%).
In addition, three-quarters (76%) of companies needing a data overhaul said that their leadership does not fully grasp how to generate value from the data the business collects.
Underdeveloped data strategies are having a real impact on businesses’ AI goals. Just 42% said that they are capable of training GenAI models on their own data, while 89% said they struggle to incorporate business data in using AI.
Among the biggest challenges deterring organisations from moving forward with AI is the risk of data exposure (39%), an inability to capture or access the right data to build models (39%), and issues around compliance (36%), highlighting problems across the board in AI governance as well as skills shortages.
Many organisations are also facing a misalignment between AI initiatives and core business objectives, with 64% of business leaders admitting their companies frequently rollout AI solutions without establishing a viable use case, while even more (74%) agreed their firms often jump into AI pilots without any clear plans to scale.
However, those organisations with greater data maturity were found to deploy AI initiatives less aimlessly. While 83% of companies that require a data strategy overhaul often or always deploy AI solutions without identifying use cases, the figure falls to 60% for data-mature companies.
SoftServe director of AI and data science Iurii Milovanov commented: “GenAI and data have a mutually beneficial relationship. Stronger data results in more robust GenAI implementations, while business-aligned GenAI applications convert unused data into valuable assets. The key is finding the balance to create a virtuous cycle.”
The report coincides with the publication of new research from Snowflake which reveals that 93% of UK businesses have achieved efficiency gains by adopting generative AI, surpassing the global average.
The “Radical ROI of Generative AI” research report, created in collaboration with Enterprise Strategy Group, surveyed 1,900 business and IT leaders across nine countries actively using generative AI for various applications.
However, despite the positive embrace of AI technologies, challenges remain over the use of unstructured data, with only 10% of UK businesses achieving high capability, compared to 20% globally. Furthermore, 63% of UK respondents cited time-consuming data management procedures as a significant hurdle, higher than the global average of 55%.
Related stories
Frenemy at the gates: Industry battles with future of AI
AI spend set to soar despite mixed results from adoption
‘AI is too often implemented, not adopted,’ study warns
Cavalier approach to data holding back adoption of AI
Come clean: ‘Dirty data’ costs UK firms £900bn a year
Get data shipshape for GenAI or risk sinking, firms told
Get to grips with data hygiene if you want to clean up
CMOs admit lack of expertise is thwarting AI adoption
Generative AI threatened by unresolved martech issues
Be the first to comment on "Firms failing to tackle data basics before rushing into AI"