|
|
|
Jul 15, 2026
|
|
|
|
|
Supported by
|
|
|
|
|
|
|
|
|
|
On Tuesday, the White House announced the launch of the first program to come out of its early June executive order on AI cybersecurity. The initiative, dubbed Gold Eagle, is a clearinghouse composed of government agencies and companies that allows them to coordinate on cyber vulnerabilities and responses. During a press briefing on Tuesday, a senior White House official lauded the contribution of open-source software providers, though they declined to name the specific companies or organizations that were involved. In response to a reporter’s question about a rumored executive order targeted at curbing Chinese open-source technology, the official stressed the White House’s backing of U.S. firms. “I could not be more clear that we are in full support of the U.S. open source community,” they said. “We will do everything we can to support the strength of that community.” Key sections of the June executive order still need to be hashed out, including a voluntary framework where AI companies can submit their models to the government for review before releasing them to trusted partners and the public. The order stipulates that the framework will be developed within 60 days of its signing on June 2.
|
|
|
|
DeepSeek’s annualized revenue reached between $400 million and $500 million recently, in a potential boost for the Chinese AI developer’s effort to raise its second funding round, The information reported. DeepSeek’s revenue growth, driven by sales of cloud-based access to its AI models, comes at a critical time. The company, which last month closed its first-ever funding round of 50 billion yuan ($7.4 billion), is already in talks to raise another 50 billion yuan in a new funding round, which values the firm at 500 billion yuan ($74 billion), The Information reported. By keeping the cost of running its models low thanks to improvements in its computing infrastructure, DeepSeek has maintained a gross margin of more than 50% from selling access to V4, its latest flagship model—despite charging for access to its models at a fraction of the cost of leading models from OpenAI and Anthropic. While working on the new funding round, DeepSeek has also hired investment banks to help it prepare an IPO in Shanghai’s tech-focused Star Market, The Information reported. The firm aims to file the listing application this year and go public next year.
|
|
|
|
IBM stock shed 25% on Tuesday, as the firm reported a shortfall in purchases of its mainframe computers resulting from customers shifting spending towards other hardware affected by AI-driven price increases. IBM said infrastructure revenue tied to sales of its mainframes, or supersized computers used by enterprises such as financial institutions to process credit card transactions, fell 7% from the previous year, more than the low single digit decline the company previously expected. IBM is scheduled to report June quarter earnings on July 22, but it gave a preview of its results on Tuesday, as companies often do when the outcome varies significantly from what they have previously projected. Overall, IBM said it plans to report revenue for the quarter rising 1% from the previous year to $17.2 billion, while earnings per share fell 2% to $2.27. The firm said software revenue rose 5% in the June quarter, a deceleration from the segment’s 11% growth during the March quarter. IBM shares hit an all-time high above $330 in June but stood around $217 as of Tuesday’s market close.
|
|
|
|
Apple has received key approval in China that paves the way for the release of its Apple Intelligence on-device AI service in the country, as the Chinese internet regulator permitted the Apple AI service, along with six other services, under generative AI rules. Apple worked with Chinese internet giant Alibaba Group in the development of its AI service for Chinese iPhone users, The Information previously reported. Other on-device AI models approved by the regulator were from Chinese smartphone makers Huawei, OPPO, vivo, Xiaomi and Nubia, as well as South Korea’s Samsung. An Alibaba spokesperson said Qwen—Alibaba’s AI model—will be integrated into Apple Intelligence across iOS, iPadOS, macOS and visionOS for users in China, giving them access to text and image understanding and generation without switching between tools. In June, Apple introduced the latest Apple Intelligence updates at Worldwide Developers Conference 2026, where it previewed a new generation of Apple Intelligence and a more capable Siri AI with new features.
|
|
|
|
Google DeepMind co-founder and CEO Demis Hassabis called for a new standards body led by the U.S. that would perform testing on frontier models. In a lengthy post on X, Hassabis wrote that the government could establish a group modeled on self-regulatory organizations or public-private partnerships like the Financial Industry Regulatory Authority. Hassabis proposed that its board include independent technical experts and open-source representatives and that funding would “mostly” come from industry. Hassabis’ post comes at a pivotal moment for AI regulation in the U.S. The Trump administration has waffled on how to create a regime that would test frontier models for safety while fostering innovation, with Trump signing a long-debated executive order in June that established a voluntary system. The Trump administration also rebranded and downsized a small agency within the Commerce Department called the Center for AI Standards and Innovation that largely fits the description of Hassabis’ proposal. Last month, the White House imposed export controls on Anthropic after cybersecurity fears over its latest public model, Fable. In his post, Hassabis wrote that the new body would initially be voluntary, though it would soon be formalized and required for labs to deploy new models in the U.S. “We need to get the technical foundations right by coordinating around a shared global framework,” Hassabis wrote.
|
|
|
|
| |
