The final trading day of October brought a sharp reassessment of AI’s financing, even as companies pushed ahead with ambitious infrastructure experiments. Here are five developments from Friday and early Saturday, October 30–31, spanning capital markets, data centers, regulation, safety and space-based computing.
1. OpenAI reportedly trims fundraising target to $20 billion
OpenAI has reduced the target for its latest funding round by $10 billion to about $20 billion, Reuters reported, citing a Financial Times account based on people familiar with the matter. The reported target valuation is $1.25 trillion. A Friday deadline for investor commitments was expected to yield a first close of less than $20 billion after some participants scaled back allocations. Reuters said a UAE investor group including state-backed MGX was expected to be the largest participant; the proposed arrangement could give OpenAI an option to lease computing capacity developed in the region. OpenAI and several named investors had not immediately commented to Reuters.
This is a reported adjustment, not a completed financing round. Its significance lies in the timing: OpenAI is seeking substantial new capital ahead of a potential 2027 listing while investors are becoming more discriminating about the cost of building and running frontier models.
2. Nvidia considers backing debt for an Anthropic-linked data center
Nvidia is in talks to guarantee billions of dollars in debt financing for Nscale’s planned AI campus in West Virginia, TechCrunch reported, citing reporting by The Information and Bloomberg. According to Nscale’s filing, Anthropic signed a $44.6 billion contract covering the first building, a planned 460-megawatt facility expected to use Nvidia’s Vera Rubin chips. A second agreement worth roughly $45 billion remains under discussion, not finalized.
The proposed residual-value guarantee would reportedly put Nvidia behind parts of the project’s construction financing if Nscale defaulted, potentially leaving the chipmaker to help find a new owner and tenant. No final guarantee has been announced. The arrangement illustrates how chip suppliers, cloud operators and model developers are becoming financially intertwined: demand for compute is substantial, but turning contracts into powered, operating facilities requires construction capital and grid access.
3. SEC scrutinizes what AI companies tell investors
The U.S. Securities and Exchange Commission is examining AI-related corporate disclosures as investors question whether complex financing arrangements and spending plans are being described clearly, Reuters reported Friday. Enforcement director David Woodcock said the agency’s principal concern is that companies disclose material information accurately so investors understand the risks. “Enforcement is a very necessary piece of making sure people are doing that,” he said.
The warning came during a tech selloff. Reuters estimated that more than $2 trillion in Nasdaq 100 market value was on course to disappear over October as investors reassessed AI spending. Monitoring disclosures is not itself an allegation of wrongdoing or an announced enforcement case; it does signal that valuation assumptions, financing exposure and the timing of promised returns may face closer examination.
4. OpenAI tightens ChatGPT’s violence-prevention safeguards for teens
OpenAI said Friday it will restrict practical information about weapons and security systems for teen users even when a request appears innocuous, Reuters reported. For users of all ages, ChatGPT may end conversations that appear fixated on violence and temporarily pause accounts after repeated such chats. The company is also limiting detail about past mass attacks and improving age prediction to place users it estimates are under 18 into a teen experience.
OpenAI said the changes will begin rolling out over the next few weeks and that it plans to notify parents when a teen account is paused for repeated discussion of violence. Those are planned controls rather than proof of their effectiveness. Their performance, especially for incorrectly classified ages and ambiguous conversations, will matter as scrutiny of chatbot safety for young people grows.
5. Google’s orbital AI prototype begins running a model in space
Google’s Suncatcher prototype satellite is communicating with Earth and running a preloaded Gemma 4 model after a launch earlier in October, TechCrunch reported. The spacecraft, built by Planet Labs, carries four Google tensor processing units but can operate only one at a time for about 20 minutes because of power constraints; it generates up to one kilowatt. Engineers first asked it for a space-themed haiku and have also used it to report orbital conditions.
The consequential test is not the poem but whether the chips can tolerate radiation and thermal stress in orbit. Google plans two more satellites in 2027 to test greater power generation and optical links between spacecraft. This remains an engineering experiment, not a commercially viable orbital data center: power, launch economics and communications must all improve before distributed AI computing in space can be practical.
Together, the stories show the central tension in AI today: laboratories and infrastructure providers continue to push technical boundaries while financiers, regulators and users demand clearer evidence of cost, accountability and safety.
