The Information AM

Save 25% on an annual subscription to read the most important news about technology and business first. For even more, Save $250 on The Information Pro for unlimited access to our proprietary org charts, databases and surveys.

OpenAI is in early talks with investors to raise a pre-IPO round and is targeting raising $30 billion, according to a person familiar with the matter.

The ChatGPT maker could seek a valuation around $1.4 trillion, though it hasn’t signed a term sheet yet with any investor, the person said.

The round, largerly driven by investors’ interest, comes before an expected IPO in 2027. OpenAI is seeing growth from selling to enterprise as its annualized revenue nears $70 billion in recent weeks, up around 70% from its revenue pace at the beginning of the third quarter. Bloomberg earlier reported on OpenAI’s fundraising plan.

President Donald Trump announced on Tuesday that leading AI companies had agreed to non-binding commitments on AI safety provisions, including internal controls on model alignment and external auditing. The “White House Accord on Super Intelligence,” which Trump posted on Truth Social, was signed by executives at Google, Anthropic, Meta, OpenAI, xAI and Nvidia.

Separately, President Trump signed an executive order formally instructing his administration to stop using the terms “artificial intelligence” and “AI” in favor or “super intelligence” and “SI.” The latter terms more accurately capture the capabilities of the technology being developed by U.S. labs, according to the order.

“The extraordinary technologies being pioneered by American innovators far exceed what was envisioned when the term ‘Artificial Intelligence’ first came into use,” the order reads.

The accord came after the White House hosted a lunch with tech leaders from those companies, along with Jeff Bezos, Satya Nadella, David Sacks and others. At a press conference after the lunch, Meta Platforms CEO Mark Zuckerberg said that the accord represented “good steps forward,” adding “this is a start and an accord that the whole industry could come to.”

The announcement comes as the House remains in recess until after the midterms, precluding the possibility of any legislation around AI safety. Meanwhile, Trump has expressed that any regulation should come from the industry itself, rather than from the government. During Tuesday’s press conference, he referred to the accord as “morally binding.” The document itself specifies that the accord may be codified into laws or regulations in the future.

Oura, the maker of a popular ring that tracks people’s health, said Tuesday it was postponing its initial public offering, “despite strong demand, due to uncertainty in the IPO market.”

Oura was scheduled to go public later this week. The company had made public its IPO filing at the beginning of this month and last week it had disclosed its tentative pricing, making clear it wanted to take the company public at a premium valuation.

But the company faced some skepticism. One analyst, Arete Research’s Richard Kramer, likened Oura to other “single product plus subscription models that didn’t end well,” such as Peloton, GoPro and Fitbit. Moreover, wariness is spreading among investors about AI financing, possibly influencing Oura’s thinking.

Oura isn’t under pressure to go public, as the company makes money and nearly all the money the company would have raised from the offering would go to satisfy tax payments related to employee stock awards.

GMI Cloud, a five-year-old provider of Nvidia chip servers to enterprises, said Wednesday it raised $668 million in equity and debt financing. The new fundraise indicates Nvidia is continuing to help smaller, upstart cloud providers, especially those that plan to boost Nvidia in more ways than one.

GMI, which stands for general machine intelligence, is one of seven cloud providers that are building facilities entirely based on Nvidia’s specifications, meaning they use Nvidia’s full stack of hardware and software. It’s also a participant in Nvidia’s AI compute partnership, a new program in which Nvidia provides credit support for new facilities in exchange for taking a cut of the eventual cloud revenues those facilities generate.

ARCHIV, a new AI and robotics investment firm, led the $223 million equity portion of the GMI funding round, in which Nvidia also participated. Taiwanese bank CTBC led the $445 million credit facility. The company did not disclose a valuation, but said it has more than $600 million in annualized revenue under contract.

The new funds will help GMI Cloud grow its AI capacity in both the U.S. and Asia. Just under 30% of GMI Cloud’s revenue currently comes from Chinese customers, but that will fall to closer to 10% next year, said CEO Alex Yeh, whose family is involved in the Taiwanese semiconductor industry. Large U.S.-based customers include OpenRouter, Fireworks AI and Higgsfield. Yeh said GMI Cloud is now processing one trillion AI inference tokens per day, 20 times higher than six months ago.

Anthropic has agreements to pay SpaceX up to $84.5 billion to use the rocket and AI company’s Nvidia-based computing capacity through 2029, according to a confidential initial public offering prospectus reported by Reuters on Tuesday. Those agreements can mostly be cancelled with a 90-day notice period, Reuters reported.

But the reported $84.5 billion total is far larger than what SpaceX has previously disclosed about Anthropic’s potential spending. In May, Anthropic announced that it had entered into an agreement with SpaceX to lease compute. SpaceX disclosed in its IPO paperwork that Anthropic had agreed to pay $1.25 billion per month in a deal that could run for three years, totaling nearly $45 billion. At the time, the company said the agreement could be cancelled with 90 days written notice.

Anthropic expects to spend at least $518 billion over the next ten years on its AI infrastructure buildout that includes six partners such as Google, Amazon, and Microsoft in addition to SpaceX, Reuters reported. The company is one of several AI labs that has agreed to lease compute from SpaceX, including Google and Reflection AI.

Google is in the process of joining Apache Ossie, a software industry group with members such as Snowflake and Nvidia that is trying to make it easier for AI tools to access data from applications and databases, according to a company spokesperson.

The search and cloud provider will be latest major software firm to join Apache Ossie, which has been working since last fall to develop standard definitions of business metrics like gross revenue and net revenue so that AI tools can better understand the context of data. Microsoft also recently became an Apache Ossie member, The Information reported earlier Tuesday.

Google joined an early version of the standards group last November but hadn’t previously revealed its intention to participate in Apache Ossie. That’s because the company was waiting for the project to officially join the Apache Software Foundation, which it did in July, said a person close to Google.

OpenAI is nearing $70 billion in annualized revenue, up around 70% from its revenue pace at the beginning of the third quarter, according to a person briefed on the numbers.

That growth comes as the ChatGPT-maker has doubled down on selling to enterprises. OpenAI has rapidly cut the prices of its models in recent months, in part thanks to optimizations it’s been able to make to running those models, and there has been growing interest from developers in its Codex coding agent.

OpenAI’s revenue pace shows it likely has narrowed the gap with Anthropic, which pulled ahead earlier this year on the success of its Claude Code and Cowork products. Anthropic’s annualized revenue pace reportedly passed $65 billion in July and some estimates have placed it at $70 billion or higher more recently. Anthropic is preparing a public offering that could take place in November. Axios first reported the details of OpenAI’s revenue pace.

Meta Platforms is adapting its AI agent Muse for small businesses, adding integrations with workplace software including Salesforce’s Slack, Intuit QuickBooks and Zoom.

In a blog post Tuesday, the company said Muse can connect to small-business owners’ professional Instagram accounts, Facebook pages and Meta ad accounts to create content and ads. Meta also highlighted other uses, including managing cash flow and inventory and communicating with customers. Additional integrations include Asana, Notion, Canva, Lovable, Box and Granola.

“Small businesses have been growing on our apps for nearly two decades,” Meta said in the blog post. “They told us they’re short on hours, not ideas. So we built Muse for Small Business to help get work done with the tools they already use.”

Meta launched its Business Agent in June, allowing businesses to automate customer interactions across WhatsApp, Messenger and Instagram, including answering questions, recommending products and booking appointments. The latest Muse integrations allows small business owners to manage internal operations and oversee accounts across multiple apps in one place.

The expansion follows Meta’s announcement Monday that it is launching a new division, Meta Enterprise Platform, to sell its AI tools to businesses as it looks to diversify its revenue beyond advertising.

Amazon introduced new AI-powered tools to help marketers buy ad inventory across the vast variety of media it sells, as it steps up its effort to strengthen its position in the ad tech world.

One of these tools, full-funnel campaigns, is designed for advertisers who already sell on Amazon’s shopping site, and allows them to buy spots on Prime Video and other types of media as well. Another ad tool, Display Video and Audio+ campaigns, helps automate ad buying across media Amazon sells, excluding its retail site. Both tools use Ads Agent, an agentic interface, to make the process of setting up the campaigns simpler.

Google and Meta have offered similar ad formats which automate ad buying across disparate types of media, which have been popular offerings for both companies.

“Previously, advertisers had to manage … separate campaign types with separate workflows and separate optimization,” said Kelly MacLean, vp of engineering, science and product at Amazon Ads in an interview.  “ DVA+ is now unifying this into a single, simple experience. So advertisers can just pick a goal… If they want to drive sales, or increase traffic, or build awareness, they can…. leverage our AI to build creative for them, set a budget and launch.”