Oracle has just appeared in a joint public letter signed by over 100 companies, warning that AI-driven cyberattacks are about to break out on a large scale. In the same week, Nvidia’s earnings report called out the Vera Rubin rack as already running on its platform. On one side is future revenue; on the other is systemic risk.

Moody’s poured cold water: AI investment will weaken the credit metrics of U.S. cloud giants. The more aggressive Oracle’s cloud infrastructure buildout is, the higher the depreciation and capital expenditures, and the uglier the short-term free cash flow will look.

The public letter made no commitments or deadlines—it reads more like a market-education document. If AI safety incidents really continue to escalate, will cloud providers’ defense costs end up wiping out profits?

The controversy: Is Oracle actually a guardian of AI safety—or the next target to be attacked?