Riverbed: AI adoption hindered by data quality, readiness concerns

Artificial intelligence, including generative AI, continues to elicit excitement about its potential across industries, but according to recent survey results, the reality doesn’t exactly live up to the hype.

AI promises to improve operational efficiencies, drive productivity, and streamline workflows, but those benefits remain dependent on several other factors, according to the Riverbed Global AI & Digital Experience Survey. For instance, while 94% of respondents believe that AI will help their organization deliver a better digital experience for end users, just 37% consider their organization to be fully prepared today to implement its AI strategy.

“Despite the enthusiasm, our study uncovered several gaps organizations must overcome to realize the full potential of AI. What leaders really want is to move from the AI hype to practical AI that works and delivers measurable results,” says Jim Gargan, chief marketing officer at Riverbed. For the Riverbed survey, Coleman Parkes Research in June 2024 polled 1,200 IT, business, and public sector decision-makers across seven countries to gauge how AI and genAI is being applied in organizations.

The research uncovered that many organizations have moved beyond the assessment and experimentation phase of AI, and today 65% are accelerating their AI strategies with investments in both infrastructure and talent. Another 23% are in the final stage in which AI is fully integrated into their business processes.

Still, organizations continue to struggle with data collection and normalization as well as their organization’s readiness to implement AI. The survey showed that 85% consider data a critical factor in implementing AI, however 69% said they were concerned about the effectiveness of their organization’s data for AI usage. Just 43% rated their data as excellent for completeness and 40% for accuracy, and 42% said their data quality is a barrier to further AI investment. Data security concerned another 76% of respondents who said they worry about their proprietary data being accessible in the public domain, Riverbed reports.

“If you have better data, that means you get better AI, you get more precise AI outcomes,” Gargan says. “More data is out on the edge, on edge devices, in the cloud, and in data centers, really spread all over the place. Organizations have to move data from where it is to where it needs to be in order to really implement generative AI. Having the ability to do that in a safe, secure, fast, and efficient way is one of the things are IT teams, in particular networking departments, are working with.”

Concerns around data lead into the readiness gap that Riverbed identified in the survey results. Just 37% of respondents say their organizations are fully prepared to implement AI projects now, and 71% say because AI is still maturing, that it continues to be challenging for their organization to implement AI that works and scales successfully. Yet on the positive side 86% of respondents said they believe their organization will be fully prepared to implement AI in three years, according to the report.

Other concerns uncovered in the survey center on the reality of AI versus the hype. A majority of respondents (82%) believe they are ahead of their competitors with AI adoption for IT services and digital experience. Only 5% of respondents said they are behind their competitors. Riverbed points out that these responses reveal a gap in perception versus reality, indicating many of those surveyed are overconfident in their AI journey.

“We are seeing now that enterprises are seeing past the inflated expectations of the AI hype, and they are beginning now to embrace a more pragmatic approach. They want practical AI that they can deliver now, that will scale and give meaningful insights. It’s time to make it real,” Gargan says.

Read more about AI readiness

• Using genAI in IT operations boosts productivity, but security concerns linger: IT managers believe generative AI can improve IT service performance, but they remain concerned about AI data quality and security, according to new research from EMA.
• Cisco debuts CCDE-AI Infrastructure certification: Cisco’s new certification aims to help prepare IT pros to design, provision and optimize networks and systems needed for demanding AI/ML workloads.
• There’s a new difficulty in tech hiring: how to recognize a good candidate: Hyped technologies create staffing pressures. And as technology gets more complex, it’s harder to understand what skills are needed and identify who has them.
• Cost, AI, and staffing are biggest concerns for data centers: Data center operators are grappling with rising costs, the impact of AI on capacity requirements, and persistent staffing shortages, Uptime Institute reports.