Research confirms AI adoption growing but governance is lagging

While it’s true that the adoption of artificial intelligence in various applications is yielding tangible results for all kinds of enterprises, there is a downside: AI’s full potential isn’t being realized because of a lack of human expertise to optimize it for business purposes.

A new global research project conducted by Juniper Networks and Wakefield Research and released June 15 shows an increase in AI adoption during the last 12 months, but a shortage of human talent is holding a great deal of good implementation back. Governance policies involving AI continue to lack maturity, the report said, and this is also a stumbling block. Both of these factors are needed to responsibly manage AI’s growth when considering privacy issues, regulation compliance, hacking and AI terrorism, the survey said. 

“IT people may know how to engineer the data, the solutions and build them so they’re secure and scalable, but we don’t necessarily have the subject matter expertise to do the actual analytics that tends to reside in the business functions,” Juniper CIO Sharon Mandell, who helped supervise the study, told VentureBeat. “I think you’re seeing this trend more and more where we want to make sure that they’re there. We want to help them with analytics tools that we can feel pretty confident are going to do the job for them. And we want to set them up so that the data is of high quality, and well-engineered so that they don’t spend all their time doing Data Prep, which is not necessarily their skill set.

“We want them to focus on analytics. As a result, if the AI is embedded in the tool, you’re putting it out at the edge of your organization. If it’s not centrally controlled, and you have to, like with many other things, maintain visibility, understand how it’s used, and understand how it’s secured. This is how you’re really getting the value that you intended.”

The survey indicates that the disparity between the substantial increase in AI implementation in the enterprise and the immaturity of AI governance and policies is staggering, Mandell said. 

“It will be critical for governance to pick up the pace so that the positives of AI deployment overshadow existing fears of whether AI can be effectively controlled. This is a challenge not unique to AI, but all emerging technologies,” Mandell said. 

A few key takeaways from the survey:

• While Juniper’s 2021 report previously showed only 6% of C-level leaders had adopted AI-powered solutions across their organizations, this year, 63% of company leaders surveyed say that they are at least “most of the way” to their planned AI adoption goals.
• IT leaders do not see AI replacing humans. Around half of IT leaders (55%) say AI will allow employees to focus on being more innovative, gain new skills (50%) and increase their engagement (47%).
• Almost all AI/ML leaders (95%) agree cybersecurity is a critical component of maintaining and securing an enterprise AI solution.
• Cybersecurity substantially increased in importance as the most critical area for AI adoption: 29% said cybersecurity was the most critical to AI adoption in 2022, versus 14% in 2021.
• Almost all IT leaders (96%) say that in the next 12 months, AI will assist in reducing risk and increasing quality within their organization, with networking/cloud (25%), IT Infrastructure (21%) and supply chain (15%) as the business functions thought to have the greatest potential to derive benefits from implementing AI.

Hiring the right people to operate and develop capabilities is a top area for investment for optimizing AI, Mandell said. IT leaders rank three areas as the top AI investment options (21% each): hiring the right people to operate and develop AI capabilities; further training the AI models; and expanding the capabilities of the current AI tool into new business units, the report said. 

“AI is ultimately taking automation to perform tasks on par with humans and many of Juniper’s own customers are leveraging cloud AI in their networks to cut support tickets, which frees up IT  teams from the drudge of tactical issues to actually focusing on improving end users’ experiences,” Juniper Chief AI Officer Bob Friday said in a media advisory. “But with all the positives, enterprises more than ever need to responsibly manage AI’s growth with proper governance to stay ahead of regulation and minimize potential negative impacts. 

“In Europe, for instance, we are seeing regulators starting to classify certain AI use cases as risky and requiring CE  certification,” he said. “AI regulation is changing quickly and business leaders must make AI governance a strategic priority.” 

The survey queried 700 senior leaders around the world with direct involvement in their organization’s AI and/or machine learning (ML) plans or deployments.