Artificial intelligence (AI) promises a revolution, but a new study by IDC and NetApp reveals up to 20% of AI projects fail due to inadequate data infrastructure.
The study explores the challenges and benefits of AI implementation at various stages of organisational maturity. A key finding is the disparity in data access across the AI maturity spectrum. Organisations categorised as “AI Masters” – those with the most experience in AI – optimise their infrastructure for easy data access. This includes a unified hybrid/multi-cloud environment and support for various data types. AI Masters report higher rates of instant data availability (48% structured, 43% unstructured) compared to less mature organisations.
“With intelligent data infrastructure in place, companies have the flexibility to access any data, anywhere with integrated data management…” said Jonsi Stefansson, Senior Vice President and Chief Technology Officer at NetApp.
“Infrastructure decisions made during the design and planning process of AI Initiatives must factor in architecture flexibility,” said Ritu Jyoti, Group Vice President, Worldwide Artificial Intelligence and Automation Research Practice, Global AI Research Lead, at IDC.
“The nature of data inputs to AI and GenAI workstreams means easy access to distributed and diverse data—both structured and unstructured data sets with varying characteristics—is critical.”
The research also emphasises the importance of data governance and security for successful AI initiatives. Only 8% of less mature organisations, categorised as “AI Emergents,” have standardised governance policies across all projects, compared to 38% of AI Masters.
The IDC report, “Scaling AI Initiatives Responsibly,” surveyed over 1,200 global decision-makers across IT, data science, engineering, and software development for AI initiatives, with 25% from Asia Pacific.