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.

Nvidia is reconsidering the structure of an initiative it announced earlier this summer, the AI Compute Partnership, to provide credit support to AI cloud providers that rent out Nvidia chips, in exchange for a cut of those rental revenues, The Information reported Monday. People involved in the discussion said Nvidia is fundamentally rethinking how the program will work.

Some more established cloud firms such as Nebius declined to participate because they didn’t want Nvidia to eat into their potential profit margins and could raise debt in other ways, the report said. Meanwhile, Nvidia has worried that some of the cloud firms that want to participate in the program would become too financially dependent on the company, said another person. Nvidia is adjusting some of the term sheets and contracts to avoid such a scenario, the report said.

OpenAI wrote to the New York City Council last week recommending that the city put safeguards in place to prevent against the risks of advanced AI, City Council speaker Julie Menin said during a press briefing on Monday.

In the letter, shared with The Information, OpenAI head of U.S. & Canada policy and partnerships head Chan Park wrote that the company wanted to work with New York City cyber defenders to provide tools, such as through its Daybreak cybersecurity program. He said that OpenAI is already working with New York state on cybersecurity efforts.

The press briefing came ahead of a public hearing on Monday in which representatives from OpenAI, Anthropic, Google and Meta are scheduled to testify in front of the city council about the risks of advanced AI. The hearing follows weeks of high-profile AI-powered breaches and warnings from AI safety researchers about the dangers of AI.

Some of these safety researchers and whistleblowers, such as former Anthropic researcher Jacob Coxon, former Google DeepMind researcher Alex Turner and former OpenAI researcher Daniel Kokotajlo, were scheduled to testify on Monday.

Menin also said that the City Council asked Elon Musk’s SpaceX’s AI unit to attend the hearing as well but the company didn’t respond. The City Council then issued a subpoena compelling it to attend. Menin said that SpaceX responded to the subpoena with a letter indicating that it wanted to work with the city, and Menin said that the city is pursuing legal action in response to SpaceX’s failure to attend the hearing. SpaceX did not respond to a request for comment.

Reflection AI, a 2-year old Nvidia-backed startup, announced its first open-weight model, adding to a collection of U.S. competitors aiming to catch up to the performance of similarly free models from China.

In its announcement Monday, Reflection said it is still evaluating and testing the long-awaited model and plans to release it later this month. Reflection aims to be the U.S. answer to Chinese companies like DeepSeek, whose AI models have dominated the open weight market. Open weight models are free to download, which offers greater security, customization and predictable pricing compared to proprietary models from companies like OpenAI and Anthropic.

Beam has about 500 billion total parameters, the numbers that make up an AI model and determine its behavior. That makes it about half the size of Thinking Machines Lab’s open weight model Inkling and around the same size as Nvidia’s Nemotron 3 Ultra, which the chip giant released in June. Reflection says that Beam outperforms both models in tests of coding and reasoning capabilities, but comes in behind leading Chinese models.

More western companies have turned to developing open weight models in recent months. Nvidia is also working on a larger Nemotron model, with assistance from a coalition of AI companies that includes Reflection.

Singapore-based DayOne Data Centers on Monday filed for an initial public offering on the Nasdaq Stock Market, a signal of optimism about the IPO market. A string of companies have postponed plans to go public in recent weeks due to choppy market conditions.

DayOne, which mainly operates data centers in Southeast Asia, reported fast-growing revenue as well as widening losses in its filing. Its revenue increased more than twofold to $484 million in 2025 from 2024, and more than threefold in the first half of this year to $512 million compared to the same period last year. The company also recorded net losses of $369 million in 2025, and $82 million in the first six months of this year.

The company was initially set up as a Singaporean subsidiary of Shanghai-based data center firm GDS Holdings, and later rebranded to DayOne after raising money from outside investors. DayOne’s top three shareholders are Coatue Management, HillHouse and GDS, each of which owns about 19% of the company.

The filing didn’t say how much Dayone intends to raise or the timeline of the IPO. Bloomberg reported in August that the company was seeking to raise $5 billion.

OpenAI said Monday that it will test a new visual ad format that uses images to show how products and services could fit into people’s lives. The ads will first appear while users generate images in ChatGPT, labeled and kept separate from the image being created. Testing begins later this month in the U.S. with an initial group of advertisers. Google and Meta already use AI to help advertisers create product and lifestyle images. OpenAI is taking a different approach, testing ads that appear alongside images people generate in ChatGPT.

The tests follow OpenAI’s notifications to some business partners that it will no longer accept advertising in ChatGPT for image- and audio-generating products, which compete with its own features. OpenAI’s new visual ad format will initially appear directly within ChatGPT’s image-generation experience.

Separately, OpenAI is expanding measurement tools and partnerships for brands advertising in ChatGPT via a string of partnerships with ad services firms. The move follows complaints by some advertisers about deficiencies in OpenAI’s advertising platform following the company’s launch of ads in February.

OpenAI said it is working with customer data companies Hightouch, Tealium and LiveRamp so advertisers can more easily send conversion data from their existing systems to ChatGPT ads. It is also partnering with marketing measurement startups Haus, Measured and WorkMagic on geo-based experiments. ChatGPT ads now also work with attribution firms including AppsFlyer, Adjust and Airbridge, which track app installs and conversions, and Triple Whale, an analytics provider for e-commerce brands.

Kling AI, a spun-off unit of Chinese video app Kuaishou Technology, has picked banks for a Hong Kong initial public offering that could raise at least $1 billion, Bloomberg reported.

Kling, which develops AI video generation models, is aiming for a listing as soon as next year. The company is working with China International Capital Corp., Goldman Sachs and UBS on the share sale, according to the report.

Kling was spun off from Hong Kong-listed Kuaishou, a major competitor to ByteDance’s Douyin video app in China. Its public listing comes as it faces a tough battle in the AI video market against Chinese and American competitors such as ByteDance, Alibaba, MiniMax, Google and xAI.

In July, Kling raised nearly $3 billion at a pre-money valuation of $15 billion. Its investors include Chinese investment firms CPE and Guofang Investment, investment bank Citic Securities and Abu Dhabi-based investment firm BlueFive, as well as Tencent, Alibaba Cloud and Baidu.

This week on AI Deep Dive

Check out the most recent episode of AI Deep Dive. An upclose look into how frontier AI actually gets built — the models, the money, and the people behind them.

Popular articles