Amid the rush to adopt generative AI, companies are still in preparation mode and are focused on building out the proper infrastructure for AI and data management, Komprise concludes in a new study of unstructured data.
Komprise, which develops software for managing unstructured data, surveyed about 300 IT leaders for its “2024 State of Unstructured Data Management” report, which it published this week.
Amid the major findings are that it’s a bit of a free-for-all when it comes to restricting access to data, Komprise says.
“At this stage in the AI game, IT has not yet enforced much restriction on data, users or tools,” the company says in the report. “Enterprises are instead focused on building out the proper infrastructure and technology stack for AI by upgrading data storage and data management technologies. Another key trend is the need to improve security and governance across GenAI and unstructured data management and more broadly by filling security skills gaps in the IT workforce, our research shows.”
Source: “Komprise 2024 State of Unstructured Data Management” report
The company found that only 13% of survey respondents say they restrict what corporate data can be used in AI, while 31% have “no restrictions for users, apps or data in AI.” It also found that 44% of survey respondents are creating AI-ready data infrastructure and 32% are building their own learning models.
To make it useful for generative AI, unstructured data, such as text and images, must be prepared. Komprise found that managing governance and security is the top data prep concern, cited by 45% of survey respondents, followed by data classification and tagging at 41%.
Keeping costs down will be one of the big challenges that companies face with unstructured data management. According to Komprise, only 30% of survey-takers report they’re increasing their budgets for AI projects. Companies are also investing in modernizing disaster recovery and backups, as well as cloud migrations, it found. Cost optimization was the number one priority for data storage cited for 2025, the company said.
Source: “Komprise 2024 State of Unstructured Data Management” report
When it comes to GenAI, enterprises have concerns. The top concern was that private data would be leaked through an LLM, cited by 27% of survey-takers, followed by violating privacy and security policies at 26%. Fourteen percent said they were concerned they would use others’ private data via LLMs, while another 14% said they were concerned about AI costs. Ten percent said a lack of transparency and provenance of models was the top concern.
“Our latest survey reveals a pivotal moment in enterprise IT as organizations grapple with the transformative potential of AI while balancing fiscal responsibility,” Komprise CEO and Co-founder Kumar Goswami said in a press release. “IT leaders will also need to factor in critical data governance and security capabilities. Managing unstructured data strategically to optimize costs and use data workflows to enrich metadata is a great place to start an AI initiative.”
You can access the full Komprise report here.
Related Items:
Five Questions as the EU AI Act Goes Into Effect
Getting the Upper Hand on the Unstructured Data Problem
Unstructured Data Growth Wearing Holes in IT Budgets
The post It’s Still Early Days for Unstructured Data Management, Komprise Says appeared first on Datanami.
0 Commentaires