Poor data management is a barrier to leveraging AI in enterprises: report

Organizations around the world recognize the importance of data management for successful AI performance, but still demonstrate low levels of data maturity, according to a recent study commissioned by Hewlett Packard Enterprises (HPE).

According to the report, only 7 percent of organizations can perform real-time data pushes and pulls to enable AI innovation and external data monetization, while only 26 percent have established data governance models and can perform advanced analytics.

Even more worrying, fewer than six in 10 respondents say their organization is fully capable of handling all the key stages of data preparation for use in AI models: accessing, storing, processing, and recovering.

“The discrepancy not only threatens to slow down the AI ​​model creation process, but also increases the likelihood that the model will deliver inaccurate insights and a negative return on investment (ROI),” said the report released on Wednesday.

The report also shows that organizations are failing to understand computing and networking requirements across the end-to-end AI lifecycle, with less than half of IT leaders admitting to understanding the requirements of different AI workloads during training. coordination and inference could be.

The ‘Architect an AI Advantage’ report, which surveyed more than 2,000 IT leaders from 14 countries, found that while the global commitment to AI sees growing investment, companies are overlooking key areas that will impact their ability to deliver successful AI results. , including low data maturity levels, potential deficiencies in their network and computing provision, and key ethical and compliance considerations.

The report states that despite the integral role of legal and compliance functions in artificial intelligence activities, 22 percent of IT leaders do not include legal teams at all in conversations about their company’s AI strategy.

“These findings clearly demonstrate the need for AI, but they also highlight very real blind spots that could stall progress if a more holistic approach is not taken,” said Sylvia Hooks, vice president of HPE Aruba Networking.

“For example, misalignment on strategy and departmental engagement can prevent organizations from leveraging critical areas of expertise, making effective and efficient decisions, and ensuring that a holistic AI roadmap congruently benefits all parts of the business” , she added.

First print: May 01, 2024 | 5:12 PM IST