AI Slowdown Call Sends Investors From Chips Toward Big Tech

Nvidia shares fell more than 3% in early trading on Monday, September 14, while Meta, Alphabet and Microsoft moved higher, exposing a sharp fault line in the artificial-intelligence trade.

The catalyst was a weekend appeal from Anthropic Chief Executive Dario Amodei, who argued that companies should slow the pace of frontier-model development as safeguards struggle to keep up with increasingly capable systems. OpenAI CEO Sam Altman endorsed the proposal, and Elon Musk, who leads xAI, posted that Amodei was right.

Investors immediately translated the safety debate into a capital-spending question. If the most advanced models take longer to train or face tougher review requirements, demand for the chips, networking gear and data-center capacity powering the race may arrive more slowly than Wall Street expects.

Nvidia took the brunt of that repricing. Other infrastructure names, including AMD, Intel, Marvell, Lam Research and ASML, also weakened, with semiconductor stocks broadly under pressure. Nvidia’s position at the center of the training boom makes it especially sensitive to any suggestion that hyperscalers or model developers might defer new deployments.

The money moved elsewhere. Alphabet gained roughly 2% in morning trading, Microsoft rose about 1.6% and Meta added around 1.4%, according to market reports. These companies are major AI spenders themselves, but they also own large software, advertising and cloud businesses that can monetize existing systems without requiring an uninterrupted escalation in training intensity.

That distinction matters. A pause in frontier research would not mean companies stop using AI. It might instead shift the market’s preference from the builders of ever-larger models to platforms selling inference, productivity tools and targeted applications.

The proposal also carries a geopolitical complication. Amodei acknowledged that China may not slow down, while President Donald Trump rejected calls for a broad pause, warning that restraint could surrender the lead in artificial intelligence. That makes a coordinated slowdown difficult, even if executives agree that safety testing needs more time.

For now, Monday’s moves look like a hedge against a slower investment cycle, not a verdict that the AI boom is over. The next test will come from corporate guidance. If Alphabet, Microsoft and Meta keep expanding data-center budgets, Nvidia’s selloff may prove temporary. If spending plans are trimmed, the rotation could deepen.