AI model shifts, chip pressure, and secure tool pivots

Techmeme reported that OpenAI and Google are providing advanced AI models to Singapore-based subsidiaries of Alibaba, Baidu, and Tencent.

The report says this arrangement exposes a gap in U.S. AI controls.

That means model access can move through regional entities even when parent-company control signals remain the headline risk.

For teams building AI products, provider choice now has to include legal entity and geography checks, not only model behavior.

Techmeme says CXMT is drawing on technology, funding, and state support as it scales.

The effort is framed around challenging top memory chipmakers rather than incremental process tweaks.

The same source says CXMT is building local supplier capacity as part of this strategy.

For AI infra teams, memory supply structure is now another architectural constraint, directly affecting model runtime and training budgets.

OpenAI is sunsetting Atlas, the AI-powered browser it launched in October with ChatGPT at its core. But it’s not giving up on the idea that AI should help people browse the web. Instead, it’s taking some of the agentic browsing features it tested in Atlas and redistributing them across ChatGPT’s desktop app and a Google Chrome extension.

The move to shut down Atlas comes a few months after OpenAI’s CEO of applications Fidji Simo told the team to cut back on “side quests,” which led to the AI firm shutting down its AI video-generation tool Sora .

For much of the past year, the AI industry had been engaged in a war to unseat Chrome as the place where people spend most of their time online. Perplexity launched Comet, The Browser Company launched Dia, and Google and Microsoft have updated Chrome and Edge, respectively, with new AI-powered features.

After a few months of experimenting, OpenAI appears to have concluded that the browser is a feature, not the destination. So it’s folding Atlas’ browser-like agent capabilities into the places people already work — and that includes Chrome.

OpenAI is launching a ChatGPT extension on Chrome that gives it access to the context of the page you’re viewing, lets users ask questions about web pages, summarize content, or start longer tasks all from the browser. It’s a direct competitor to Google’s Gemini Side Panel, which performs several of the same tasks.

Windows 11 updates could soon include fixes for more security issues at once. Microsoft said in a blog post on Thursday that it’s now using AI to “identify potential issues earlier,” which means “customers will see a higher volume of security updates included in each security release.”

Hackers, even amateurs, have increasingly been using AI to quickly exploit security weaknesses over the past several months. Security researchers are also using AI to find issues faster, leading to more frequent high-severity vulnerabilities, like the “Copy Fail” exploit that impacted nearly every Linux distribution in May.

Microsoft says it’s updating its Secure Development Lifecycle to make sure it “explicitly accounts for potential AI-enabled attack techniques and exploit paths.” It’s also making investments to “ensure that we are not compromising update quality as we gain speed,” including integrating AI more throughout its security updates process.

Microsoft emphasized that while AI will now be more involved in identifying and resolving security issues, developers will still verify those findings and “make risk-based decisions” about updates.

The Verge The Verge logo. Sign in to see your notifications or create an account to join the conversation.

Launch HN listed Context.dev (YC S26) as an API to get structured data from websites.

The post attracted 96 upvotes and 67 comments from the HN community.

The proposal targets repetitive web data extraction work that is typically handled with one-off scripts.

Developers can potentially standardize content extraction if the API output is stable and typed.