AI

The SpaceX-owned AI coding tool says the modified version of Claude Code appeared on its extension marketplace from 5:27AM ET and 4:35PM ET on Thursday. The extension contains malicious code that collects various information, including hostname, username, operating system, home directory path, public IP address, and more.

Android Authority spotted code in a beta version of the Android Google app (v17.71.17) referencing the new mode, with the tagline, “Use Gemini Live without an internet connection.” They say the code doesn’t hint at how the local-only Gemini Live mode would work, though, or whether it would support camera and screen-sharing features.

As part of the acquisition, which was disclosed under the EU’s Digital Markets Act, Apple will “make employment offers to and hire certain employees” at Ascend.io, a startup that specializes in agentic data engineering. An update on Ascend.io’s website says the startup is “winding down operations as we prepare for our next great adventure.”

The music streaming platform is reportedly aiming to release the feature, which would allow users to prompt AI-generated remixes and covers of participating songs, in the first quarter ​of 2027. This is expected to be a paid add-on for Premium subscribers, with Spotify considering charging an additional $4 to $5 a month.

OpenAI’s latest claimed breakthrough, one of three major results announced Monday, refutes the long-held Exponential Time Hypothesis. Wired captures the excitement and uncertainty among researchers. It’s a familiar story in mathematics, where a barrage of AI-generated discoveries has left researchers scrambling to understand what’s actually been achieved.

The company launched Jules in December 2024, positioning it as a tool that can complete coding jobs “you don’t want to do.” But now, Jules is shutting down, and Google says it will bring its learnings from the tool into its AI coding platform, Antigravity:

The judge in Apple’s trade secrets lawsuit against OpenAI has set a trial start date of July 19th, 2027, as reported by 9to5Mac.

However, the company has seen “promising results” with its ad experiments, chief business officer Philipp Schindler said on yesterday’s earnings call.

After making Large 4 available in a public preview earlier this month, Mistral is now releasing an improved version to everyone, which it says ranks as the “best open-weight model for coding.” The model offers advancements in cybersecurity-related tasks, like creating patches or reproducing vulnerabilities, and is more likely to refuse “malicious cyber requests” compared to other open models, according to Mistral.

Reuters reports that Finnish police said the criminal probe will look into “ambiguities and shortcomings in the drainage and water management at the construction site.” Google also reportedly cleared forests around the data center site near Muhos, Finland before completing required environmental impact assessments.

These alerts, which Meta first brought to Instagram in February, also apply to conversations about self-harm with its AI chatbot across Facebook, Messenger, and Instagram.

The New York Times reports that US intelligence officials believe an executive at one of Nvidia’s data center customers, Megaspeed, may be a “Chinese intelligence asset,” and has a “close relationship” with Nvidia CEO Huang. Details are sparse in the article, especially on how close that relationship was — Nvidia says she’s just one of “thousands of customer executives” Huang meets every year.

Developer Maxon is teasing a new AI tool generating a model based on an uploaded image, alongside options to reference text and sketches. It’s not clear how the image and sketch options will differ, but 3D creatives are reacting in the comments with other questions: Will it let you turn anyone’s art into a model?

Epic’s VP of development, Marcus Wassmer, told This Week in Videogames that “we are definitely using AI to develop code and features for Unreal Engine.” Wassmer says that the UE team has seen benefits like “faster” development of features, and the ability for engineers to “explore ideas” that they haven’t had “the time to tackle.”

AI data center projects are continuing to pop up across the US, with frequent opposition from locals concerned about their impact. Here are a few recent articles about the projects:

The company launched a new Gemini agent for enterprise earlier this month, and Google will bring that “same experience” to consumers later this year, starting with the Gemini app, CEO Sundar Pichai said on today’s Q3 2026 earnings call.

The US Department of Justice has “made inquiries” about the $13 billion deal, including how it could affect competition in the AI industry, according to a report from Reuters. It adds to existing regulatory scrutiny surrounding Nvidia, as it’s also reportedly facing a DOJ investigation over its AI chip dominance.

As reported by Semafor, the hosts of the popular podcast haven’t been using President Trump’s preferred “super intelligence” word for AI as much as they did at the beginning of the month.

OpenAI had promised this was coming for its AI agents, but now the feature is here. You can text through iMessage or RCS, OpenAI says. It’s rolling out in preview for US Pro users.

Constellation Energy will receive the funds to expand capacity at eight nuclear plants by 1.1 gigawatts. Most, if not all, of that additional capacity has already been claimed by Google, Meta, or Amazon for their data centers, according to TechCrunch.

With the new “Connected Apps” platform, developers can submit apps for people to use within Gemini and Search. This will allow developers to launch integrations similar to the ones OpenTable and Instacart built into Google’s AI tools.

Along with support for projects, which lets you assign Copilot a task across multiple repos, the GitHub Copilot app now allows you to pick up your work where you left off when you switch to other devices. You can now create reusable AI automations within the GitHub Copilot app as well.

OpenAI is the latest tech company (and according to Politico, the first major AI company) to set up a command center for fighting election misinfo and abuse of its systems, approaching the second major AI-era US election: