AI warnings by Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman could hammer this momentum stock

AI Development Pace Concerns from Anthropic and OpenAI Executives Draw Attention to CoreWeave Exposure

Bernstein analyst Madison Rezaei stated in a note that CoreWeave stands as the most exposed company in the firm’s coverage to any slowdown in AI training demand. Rezaei estimated that 25 percent of the company’s existing active U.S. power capacity, along with roughly 74 percent of its contracted power, is located in Tier 3 and Tier 4 markets. The analyst added that the backlog consists primarily of take-or-pay contracts, which limits immediate risk, but noted potential for reduced demand for additional contracted rural power if development activity moderates.

CoreWeave shares declined nearly 7 percent in pre-market trading, among the larger moves in AI-related equities that session.

Anthropic chief executive Dario Amodei addressed AI risk management in an essay published over the weekend. He wrote that fully addressing risks requires additional prudence beyond investment in safeguards, including pacing the rate of capabilities advancement so that risk prevention measures can remain aligned. Amodei stated that the pace at which AI model capabilities improve should be slowed, while noting that progress would still appear rapid and that the gained time should be used wisely.

OpenAI chief executive Sam Altman posted agreement with Amodei’s position. In a follow-up post the next day, Altman described two scenarios to avoid. The first involves loss of control over AI systems, which he said requires alignment and safety techniques to advance ahead of model capabilities. The second involves excessive concentration of power, such as one country or one laboratory gaining disproportionate influence through an advanced system. Altman characterized the desired outcome as a narrow middle path that prevents either outcome.

CoreWeave has financed expansion through substantial debt on the assumption that demand for its computing resources will remain elevated over multiple years. Any sustained reduction in the pace of large-scale AI training could affect utilization of planned capacity additions, according to the Bernstein analysis.

Market participants have monitored statements from major AI laboratory leaders for indications of shifts in development timelines. The recent comments from Amodei and Altman focused on internal risk management practices rather than specific regulatory proposals. CoreWeave’s business model centers on supplying specialized computing infrastructure under multi-year agreements, with a portion of its power procurement tied to locations that may face different demand dynamics if training workloads contract.

No direct causal link between the executive statements and the observed share price movement has been established in available trading data. Pre-market volume and broader sector movements also contributed to price action across several AI infrastructure names during the same period.