Many technology companies are rapidly developing AI infrastructure as they compete for dominance in this red-hot market. Artificial intelligence requires a lot of energy and its energy needs are expected to increase in the coming years. But this also means that there is an investment opportunity in the utilities sector. Aaron Dunn of Morgan Stanley Investment Management says the “next big bottleneck” for hyperscalers – who do much of the cloud computing for AI applications – is energy or fibre. That’s because large language models require a lot of data center capacity, he noted. “In my opinion the highlight could be the next bottleneck [to] the growth of artificial intelligence and the growth of some sort of data center and cloud computing environment,” the portfolio manager told CNBC’s “Street Signs Asia.” Dunn manages the Morgan Stanley US Value Fund. And that’s why Dunn is “pretty bullish” on utilities, naming one stock to play for: CMS Energy. “I think once data centers start really trying to take power off the grid, there will be a lot of utilities, both municipal and public, both of which will have to serve retail customers, [and now] a big load of industrial power, it’s going to need maintenance,” he said. “So from my perspective, it’s going to take a couple of years to probably be very short on electricity,” added Dunn, who is also the CMS co-lead is a company that “really promotes renewable energy” — an action in line with many hyperscalers who are “very focused” on green energy, Dunn said. They want to reduce their carbon footprint and do they need renewable energy, and so this sector is poised to grow “dramatically” in terms of capacity in the U.S., he added. “And so these utilities … have a very favorable opportunity to achieve solid earnings growth and good returns for them,” Dunn concluded. In notes sent to CNBC, Dunn also named another stock: Emerson Electric. Dunn is also a portfolio manager of the Eaton Vance Focused Value Opportunities Fund. Since 2014, the fund has outperformed the its benchmark in five of the last nine According to BofA estimates, Nvidia-supplied AI servers alone have consumed about the same amount of electricity as 20 million U.S. homes. According to BofA, data centers, which house large amounts of computing power needed for AI workloads, use between 1% and 2% of global electricity. Their energy consumption is expected to grow at a compound annual growth rate of 11% until 2030, the bank said. — CNBC’s Pia Singh contributed to this report.