One of the leading emerging use cases for artificial intelligence could be automating customer support requests, according to Barclays. At a time when some investors are beginning to doubt the AI momentum, some companies are now experimenting with AI for customer support and consumer-facing features. Barclays analysts estimate that for companies in the gig economy, AI could help drive more than 200 basis points of margin improvement from automating requests. The cost per human interaction for completed service requests is $6, while the cost per AI interaction is as low as $1 — and could fall even lower in the future, Barclays found. “These companies have complex payments, low margins, lots of customer service headaches…and hence have outsized ‘operations and support’ expense as a percentage of company [earnings before interest, taxes, depreciation and amortization],” analyst Ross Sandler wrote in a Tuesday note. He noted that Klarna announced in late February that it shifted two-thirds of its customer service requests to an OpenAI-powered assistant, a move the fintech company says could drive a $40 million profit improvement this year. He mentioned names that stand to benefit the most from using AI to handle Level 2 and Level 3 customer service requests, including Lyft , Instacart and Roblox . These support levels involve escalations from solving simple customer service problems, and can entail trouble-shooting with support technicians and the involvement of engineers, if needed. Automating customer service requests could also be an AI revenue tailwind for hyperscaler tech services, such as Microsoft’s Azure, Amazon Web Services and Google Cloud Platform, Sandler pointed out. Lyft stands to be the biggest beneficiary of AI automating customer support requests, assuming that it can move just over a third of its Level 2 and 3 human-based customer service requests to AI. Barclays estimates that AI could drive EBITDA growth up by around 8% for the ride-sharing company, which has significantly high customer service request volumes. The stock, which is up roughly 18% so far this year and more than 100% over the past 12 months, could decline about 10% from current levels, according to analysts surveyed by FactSet. Internet domain registry GoDaddy could see about 4% EBITDA growth by using AI automation in operations and support, meanwhile, according to Barclays. At its investor day on March 6, GoDaddy said it expects between 6% to 8% annual revenue growth between 2024 to 2026, fueled by acceleration in its applications and commerce business. Shares are up about 10% year to date, and analysts polled by FactSet see upside of 9%. Finally, online gaming platform Roblox also stands to benefit from the technology as it runs high frequency micro-transactions, Sandler noted. Analysts’ consensus price target implies the stock could gain more than 20%, per FactSet. Roblox’s fourth-quarter results came in better than what the Street had hoped for in February. Quarterly bookings hit $1.13 billion, its highest ever. Shares are down more than 10% in 2024, however. To be sure, it’s still early days for companies that could benefit from deploying AI as part of their customer support, Barclays found. Most of these companies are likely only using AI for a small percentage of Level 1 requests, Sandler noted, adding that many Level 2 and 3 requests are complicated and may not satisfy customers, which are expected to churn. “We’ve witnessed productivity increases across software development teams in consumer internet, but this kind of soft improvement is harder to quantify in terms of direct P+L benefit,” Sandler said.