Morgan Stanley co-president Dan Simkowitz said slower releases of frontier artificial intelligence models by OpenAI and Anthropic would not weaken demand for AI infrastructure or the financing supporting it, while investor Steve Eisman called runaway-AI warnings garbage and argued that weakness at either lab could damage Nvidia, hyperscalers and the broader trade. Anthropic is preparing an IPO that could raise as much as $100 billion at a valuation of about $2 trillion, compared with nearly $965 billion in May. Nvidia is considering investing up to $10 billion as an anchor investor. Eisman said he has stopped adding to AI positions and trimmed exposure, citing open-weight models, limited competitive moats and risks to OpenAI and Anthropic. Simkowitz said spending on computing capacity and chips remains intact, with demand from companies and consumers, and said Morgan Stanley is expanding AI use across research, cybersecurity and customer service. Prediction markets put Anthropic's chance of listing by December 31 at 87%, while a separate market assigned roughly a 73% chance of a first-day market capitalization of at least $2 trillion.