Featured Story
Intel and telcos left in virtual RAN limbo by rise of AI RAN
A multitude of general-purpose and specialist silicon options now confronts the world's 5G community, while Intel's future in telecom remains uncertain.
Enterprises are taking a more measured approach to deploying GenAI due to concerns around potential risks the technology poses and a desire for improved data governance.
Nearly half of enterprises are investing in generative AI (GenAI), but they're cautious about the rate at which they're deploying the technology, according to a report by research firm EY. While 43% of the 1,405 enterprises surveyed are investing in GenAI, 38% are moving forward with a "measured, incremental approach."
Taking a careful approach to deploying GenAI
Enterprises are taking a more measured approach to deploying GenAI due to concerns around potential risks the technology poses and a desire for improved data governance – 46% of respondents said there's a need for improved data governance to address risks around data accuracy and ethics. Enterprises are also analyzing the value of GenAI and which vendor to rely on to deploy their AI applications, said Tom Loozen, global telecommunications leader for EY.
Another factor slowing down GenAI deployments is a knowledge gap within enterprises around how to use AI, added Loozen. EY's survey revealed that 73% of enterprises want to better understand GenAI concepts and use cases.
Enterprises are also weighing the cost of deploying AI applications, determining which solution to use from which vendor, and trying to gauge how their data will work across different vendor systems, said Loozen. Organizations are facing "many unknowns" in terms of the best technology, strategy and partners to work with when deploying GenAI use cases, he said.
AI amplifies human biases
Last fall, AI thought leaders examined the challenges behind developing responsible and ethical AI during a press conference at the VMware Explore event. Human biases can be amplified in AI, which can be problematic when training AI models, explained Meredith Broussard, NYU professor and author.
"All the biases of the real world are reflected in the data we're using to train a model," said Broussard. "And therefore, the model is also going to reproduce certain biases that are created."
There's also a lack of industry standards around best practices for deploying AI applications with customers, said Richard Munro, director, office of the CTO for VMware.
Top GenAI use cases
Among the top GenAI use cases enterprises are focused on are employee training (36%) and customer-facing capabilities, such as sales, service and support (35%), according to the new EY report. Software development and testing was the third most popular choice as the most significant AI and genAI application for enterprises at 34%.
Source: EY
"As you would expect, [GenAI] is really being tried out in many different places and it has the potential of driving a lot of value in each of those places," said Loozen.
However, respondents were equally interested in GenAI use cases for product and service design (33%), supply chain management and orchestration (33%), security and fraud management (33%), and content creation and management (33%).
Combining AI with 5G and IoT
In addition to GenAI, the EY report examined enterprise interest in 5G and IoT, among other emerging technologies. About 42% of enterprises said examining how 5G is utilized together with other emerging technologies, such as AI, is among their most important 5G priorities for the future.
"What stands out through the whole research is the idea of 'how do these different technologies – 5G, cloud, edge cloud, RPA, AI – how do they come together and how do you make them work together?'" said EY's Loozen. Vendors have an opportunity to demonstrate to customers the value of these technologies and how they work together, he added.
You May Also Like