Goldman Sachs is weighing the upside and the danger of Big Tech's enormous bet on artificial intelligence, and its analysts have landed on a split verdict: there are clear winners to buy, but also real reasons for caution.

According to Benzinga, Goldman Sachs warns that Big Tech's $770 billion AI spending frenzy could backfire. That figure captures just how much money the largest technology companies are pouring into the data centers, chips, and infrastructure needed to build and run AI systems.

One consequence is already showing up in how these companies handle their cash. According to The Mercury News, Big Tech stock buybacks are vanishing as the AI spending spree eats up cash. Buybacks — when a company repurchases its own shares — have long been a way for cash-rich tech giants to reward investors. As more money flows into AI, there is less left over for that.

Goldman is not entirely bearish, though. According to Business Insider, the bank says investors should buy 16 tech stocks positioned to win as an AI-capex "supercycle" juices earnings. In other words, the same wave of heavy capital spending (capex) that pressures cash flows could also lift the profits of companies selling the picks and shovels of the AI build-out.

The through-line across all three reports is tension: spending that is large enough to reshape balance sheets, reward some companies, and potentially disappoint if the returns do not materialize.

Why it matters: Big Tech dominates major stock indexes, so whether this historic AI spending pays off — or backfires — affects the retirement accounts and investments of ordinary people, not just Silicon Valley.