Goldman Sachs chief U.S. equity strategist Ben Snider said adjusted second-quarter EPS growth for S&P 500 companies reached about 31% year on year after excluding other income such as some private investment gains, with AI infrastructure-related stocks contributing nearly half of the increase and helping keep the index near record highs. The bank said the clearest AI profits remain concentrated in chips, memory, servers, cloud capital expenditure and power equipment, while productivity benefits from corporate AI adoption have yet to show up broadly in reported results: about 11% of S&P 500 companies quantified AI-driven improvements in functions such as coding or customer support, and about 2% directly linked AI to better earnings. Goldman also said companies disclosing AI productivity gains still did not post meaningfully stronger results than peers, with median earnings growth of 17% versus 14%, a gap it said was not statistically significant; AI-related spending remains below 0.5% of S&P 500 revenue but is accelerating and could eventually shift investor attention toward AI adopters.