Safety stays the highest concern, each as a barrier to deployment and a main use case for AI itself. “It’s important to struggle AI with AI from a community safety perspective,” says Brandon Butler, senior analysis supervisor at IDC. “There’s a realization that nefarious actors are leveraging AI themselves.”
Integration with current programs and a scarcity of expert expertise comply with shut behind. “Most folk don’t really feel their employees can totally consider and choose the correct options,” Leary says. Consequently, many organizations are turning outward for assist: 81% say they’re growing spending on managed service suppliers (MSP) to help AI initiatives.
There are a selection of network-related points which are driving enterprises to delay or abandon their AI tasks, in line with IDC’s particular report on AI in networking.
IDC
Infrastructure demand is accelerating
Whilst adoption slows, AI’s impression on infrastructure is right here. “The stress is already on the community,” Butler says. “The query now could be whether or not organizations can sustain with what AI is demanding of their infrastructure.”
For example, 89% of knowledge facilities anticipate to extend bandwidth by at the least 11% throughout the subsequent 12 months, pushed by AI workloads. That demand extends past particular person services, with 91% anticipating related progress in inter-data middle connectivity, highlighting the pressure on distributed architectures.
Cloud environments are seeing sharper will increase. Organizations anticipate a mean 49% rise in bandwidth for cloud connectivity over the subsequent 12 months. “The cloud is sort of at all times concerned,” Leary says. “The most important group mixes one cloud platform with a number of information facilities.”
Edge deployments set the subsequent wave
Past the info middle and cloud, the community edge is rising as the subsequent main progress space. Right now, 27% of organizations have deployed AI workloads on the edge, and 54% plan to take action inside two years. “People who’re leveraging AI extra extensively are already pushing workloads to the sting,” Butler says. “We see this as a number one indicator of the place the market goes.”
