Dive Brief:
- The CIO is the executive most commonly tasked with overseeing governance in companies with ongoing or near-term generative AI investments, according to an Info-Tech Research Group survey of 848 IT pros and tech leaders published Wednesday.
- The second-most frequent governance structure over the nascent technology is a committee or group, as reported by 17% of AI adopters.
- One in 5 organizations adopting AI says no one is responsible for the technology yet, while 1 in 10 businesses say AI governance falls under two or more executives' purview.
Dive Insight:
Best practices for enterprise adoption of generative AI are still emerging. Many businesses are experimenting with the novel technology, creating rules along the way.
“We’re sitting here at the end of 2023, and I think it’s okay that not every organization has figured out their AI governance approach,” Brian Jackson, research director in the CIO practice at Info-Tech Research Group, told CIO Dive.
However, there’s a difference between experimenting with guardrails and letting the tech loose within workflows, Jackson said.
Nearly half of organizations want to adopt generative AI features from popular vendors in beta or when generally available, but more than one-third of executives say they need more information before making a decision. Nearly 1 in 5 are waiting until they see results from other organizations, according to the report.
“You can’t be, on one hand, saying that you’re going to be using AI for strategic-level capabilities, and on the other hand, not doing all the right things to make sure that governance is in place,” Jackson said.
The disconnect occurring in some organizations is driven by “a combination of being new to the technology and not having necessarily the right personnel or the right experience on hand,” according to Jackson.
Most of the concerns CIOs had were related to data privacy and security as organizations experiment with using, customizing and training generative AI, Jackson said.
“It’s not just like the cloud questions we previously had,” Jackson said. “It’s all of those questions and new questions about, ‘When could my data be exposed just enough to be used to train a model of a third party,’ and ‘When am I comfortable with using my intellectual property or sensitive data to get that competitive advantage?’”
Enterprise adoption of generative AI is within reach as pressure to implement the technology comes from those atop the leadership ranks down to lower-level positions.
More than 4 in 5 enterprises will adopt generative AI tools and services by 2026, Gartner predicted Wednesday. It’s an astronomical jump from less than 5% of organizations currently using generative AI tools this year, according to Gartner.
