|
|
|
May 12, 2026
|
|
|
|
|
Supported by
|
|
|
|
|
|
|
Happy Tuesday! OpenAI launches its $10 billion private-equity joint venture. Google says hackers are using AI models to find a previously undiscovered security flaw. Ilya Sutskever, OpenAI's co-founder and former chief scientist, says his stake in the firm is worth $7 billion.
|
|
|
|
OpenAI on Monday announced its $10 billion private-equity joint venture called the OpenAI Deployment Company, which aims to help businesses deploy and integrate AI into their operations. As part of the new company, OpenAI also announced it had bought Tomoro, an AI consulting and engineering firm, whose acquisition would also bring on around 150 forward deployed engineers to the new joint venture. The terms of the deal were not disclosed. The joint venture was done in partnership with 19 investment firms, consultancies and
system integrators, including TPG, Advent, Bain Capital and Brookfield, OpenAI said. It is launching with more than $4 billion in capital raised from those partners, the company said. The plans are part of a broader push by AI companies to increase spending from corporations on AI. Anthropic earlier this month announced it had set up its own joint venture, first reported by The Information, with PE firms including Blackstone, Heller & Friedman and Goldman Sachs.
|
|
|
|
Google recently observed hackers using AI models to find a previously undiscovered security flaw, the company said on Monday. The hacker planned to use the flaw for a “mass exploitation event,” but Google security researchers believe they successfully warded off the attack after noticing the hackers’ activity, the company said. The incident is one of the first confirmed cases of the type of potential cyberattack that AI labs like Anthropic and OpenAI have been warning against in recent months. Both of those labs have declined to release the latest versions of their models widely because the AI is so good at quickly finding undiscovered vulnerabilities in popular software. Instead, Anthropic and OpenAI have said they’re only sharing the models with a small number of companies and government agencies to give them time to find the flaws and patch them on their own.
Meanwhile, security experts have warned that even older versions of AI models are still concerningly good at finding and exploiting vulnerabilities. Google did not offer details on the hackers it observed finding the vulnerabilities, nor did it say which potential victims they were targeting or which AI models were used. But the company said it believes its own Gemini AI model was not used in the potential attack, and noted that it has separately observed hackers linked to China and North Korea using AI models to find security flaws. Anthropic similarly disclosed last year that it cut off suspected Chinese hackers after noticing them trying to use its models for cyberattacks.
|
|
|
|
Ilya Sutskever, the OpenAI co-founder and former chief scientist who left two years ago, owns a stake now worth close to $7 billion, he testified on Monday. The figure emerged during the trial in Elon Musk’s lawsuit against OpenAI and Microsoft. A lawyer asked Sutskever to confirm that his stake in OpenAI was worth $5 billion as of last November, and has now increased to as much as $7 billion. Sutskever said that was roughly correct. OpenAI in March completed a new financing round that raised $122
billion and increased its valuation to $852 billion, including the new money. Sutskever, a revered AI researcher, was one of OpenAI’s earliest employees. He left in May 2024 and founded Safe Superintelligence, which was valued last year at $32 billion. The trial has elicited figures for the holdings of other OpenAI stakeholders, including co-founder and President Greg Brockman, who said last week his stake is worth close to $30 billion.
|
|
|
|
Chinese social media giant Kuaishou Technology said in a regulatory filing on Tuesday that the company’s board of directors is assessing a restructuring of its Kling AI video business that may involve raising outside capital for the unit, confirming an earlier report by The Information. The Information on Monday reported that Hong Kong-listed Kuaishou, a rival to ByteDance in the short video space in China, is planning to spin off Kling with the aim of taking the AI video unit public next year, in an attempt to capitalize on investors’ appetite for
AI stocks. Kuaishou has held talks with potential investors for a pre-IPO funding round that values Kling at $20 billion, The Information reported. Kuaishou shares soared as much as 11% Tuesday morning in Hong Kong following The Information’s report, before giving back the gains and closing 1.9% higher. In the filing, Kuaishou said that its board is assessing the proposal to restructure Kling “with a view to further leveraging external financing resources.” It also said that the talks are at a preliminary stage and it hasn’t entered into any definitive agreements. Kling develops AI video generation models that compete with those from ByteDance, Google and Alibaba. Kuaishou has a market capitalization of around $30 billion. Its plan to spin off Kling is the latest example of how Chinese tech firms are
trying to take advantage of the AI boom in capital markets. Chinese large language model developers Zhipu and Minimax have seen their stock prices soar since the listings in Hong Kong early this year.
|
|
|
|
Uber promoted its Chief Marketing Officer Jill Hazelbaker to president and chief corporate affairs officer, adding human resources, real estate and safety to her remit. Hazelbaker is a veteran technology communications and public policy leader who worked at Google and Snap before joining Uber in 2015. She got her feet wet at the company by leading communications during the turbulent resignation of former CEO Travis Kalanick. Under CEO Dara Khosrowshahi, Hazelbaker has pushed the company to invest more in safety, launched numerous marketing campaigns and started to reposition the company for the rise of autonomous vehicles. Uber’s current chief people officer Nikki Krishnamurthy will be stepping down. Hazelbaker will work alongside Andrew Macdonald, Uber’s
president and chief operating officer. In an email to staff, Hazelbaker praised Uber as “one of the most complex, real-time, people-powered systems in the world.” “We’re at an important inflection point—one that demands speed, rigor, and clarity,” she added.
|
|
|
|
| |
