Data issues standing in the way of firms being 'AI-ready'

Further evidence has emerged that companies are still struggling to get their data “AI-ready”, with more than eight in ten admitting that their data sits in silos across their organisation, while even more have no real-time data accuracy system.

That is according to the Cisco AI Readiness Index, based on a double-blind survey of 8,161 senior business leaders at organisations with 500 or more employees with responsibility for AI integration and deployment within their organisations.

The report explains that, with a broad set of data sitting across different domains, applications and workloads, organisations face a real – and increasingly urgent – need to prioritize a strong data strategy to unlock the true potential of AI, despite nearly all (97%) respondents saying the urgency to deploy AI-powered technologies has increased in their company over the past six months.

Cisco states: “Beyond ensuring the security and protection of data, organisations must ensure a centrally managed database, integrated AI systems, data hygiene practices, and the proficiency of current analytics tools and their employees. This challenge is only growing to become bigger as IDC forecasted that data creation and replication will grow at a CAGR of 23% from 2020 to 2025.”

However, business leaders are increasingly recognising that effective data analytics tools go hand in hand with AI applications and overall data strategy, with more than two-thirds of global respondents (67%) positively rating the ability of their analytics tools to handle complex AI-related data-sets.

Even so, organisations are facing a challenge in the fact that 74% of respondents said their analytics tools are not fully integrated with data sources and AI platforms being used. In fact, almost one-third (31%) of respondents said their tools were not integrated or somewhat integrated at best.

While Cisco maintains that AI will transform the business landscape, it concedes that the technology also raises new issues around quality of data as many organisations are using external data to train their AI models.

Positively, companies are translating this new challenge into action. Around three quarters (76%) of global respondents are conducting advanced and intermediate level external data quality checks, to ensure the reliability of it for AI training.

While the vast majority (97%) are tracking the origins and lineage of data used in AI models, true effectiveness still varies. Only four out of 10 respondents have a structured system for tracking data origins and those systems are not integrated with all AI projects. Meanwhile, 17% of global respondents have basic tracking in place but lack comprehensive lineage details.

The study points out that in a rapidly evolving landscape, ensuring the accuracy and reliability of data is critical. Just under one-third of businesses today boast a continuous data accuracy validation system integrated with real-time corrections, indicating that more can be done by other organisations.

Maintaining data accuracy not only prioritises data security, but it also enables companies to leverage up-to-date data to make accurate decisions.

The report concludes: “Advanced data capabilities are critical to ensure that companies can leverage the full potential of AI and AI-powered technologies. However, our research reveals significant gaps in how organizations are managing data today.

“As mentioned above, the majority of respondents say their data exists in silos. In addition, a mere 21% of companies say their network has ‘optimal’ latency to support demanding AI workloads. This highlights that most companies are still input/output poor and lack basic data management capabilities.”

Earlier this week, The Software Bureau published research which revealed that so-called “dirty data” is costing the UK economy £900bn a year, with retailers the worst culprits.

Meanwhile, The State of Data and Analytics Report by Salesforce and Tableau showed more than three quarters (77%) of business leaders fear they are already missing out on the benefits of GenAI, and marketers are the most desperate, with 88% suffering FOMO.