2026-10-31 · SaturdayPublished daily at 08:00
AI Daily
Cursor discloses malware incident involving the Claude Code VS Code extension, affecting up to 5,000 users

Cursor discovered that the Claude Code VS Code extension distributed through its IDE extension marketplace had been modified and implanted with malware, potentially affecting up to 5,000 users. The malicious version was not developed by Anthropic. Cursor took the extension marketplace offline that afternoon, worked with the relevant provider to remove the malware, and restored it after confirming there was no further risk. Affected users are advised to follow the blog's instructions for checking their systems and upgrade to Claude Code for VS Code 2.1.314.
- 1Alibaba: Users of Apple Intelligence in mainland China can choose whether to enable Qwen integrationIT Home · Artificial Intelligence
- 2
- 3An Epoch AI experiment shows that GLM 5.3 can autonomously attack Firefox during the patch windowEpoch AI:Gradient Updates
01Product Releases/Updates
Anthropic renames the Claude Max plans as Max and Ultra; verified Ultra accounts are no longer subject to the five-hour limit
Anthropic renamed its Claude subscription plans: Max 5x to Max, and Max 20x to Ultra, with prices unchanged (Max $100/month, Ultra $200/month). Effective immediately, the five-hour limit has been removed for most existing subscribers’ verified Ultra accounts. New accounts retain a temporary limit to prevent fraud and abuse, while the weekly limit remains unchanged.
Original textGemini API's Computer Use tool officially becomes available (GA)
On October 30, Google announced that the Gemini API's Computer Use tool was officially available (GA), covering the Interactions and GenerateContent APIs.
Original text02Industry News
Cursor discloses malware incident involving the Claude Code VS Code extension, affecting up to 5,000 users
Cursor discovered that the Claude Code VS Code extension distributed through its IDE extension marketplace had been modified and implanted with malware, potentially affecting up to 5,000 users. The malicious version was not developed by Anthropic. Cursor took the extension marketplace offline that afternoon, worked with the relevant provider to remove the malware, and restored it after confirming there was no further risk. Affected users are advised to follow the blog's instructions for checking their systems and upgrade to Claude Code for VS Code 2.1.314.
Original textAlibaba: Users of Apple Intelligence in mainland China can choose whether to enable Qwen integration
Alibaba released an announcement explaining that, after Qwen is integrated with Apple Intelligence, users can choose whether to enable it. The feature will open for testing with the iOS developer beta on November 4 and will be rolled out to users in China when iOS 27.3 is officially released in December.
Original textProject Suncatcher runs TPUs on orbiting satellites; the Gemma model performs inference in space for the first time and sends back a response
Philipp Schmid relayed a message from Sundar Pichai: After Google Project Suncatcher launched prototype satellites several weeks ago, it successfully sent commands to an orbiting satellite, powered on a TPU, ran a prompt, and received the first complete response on the ground (a haiku); the author said Gemma 4 became the first large language model to run in space.
Original textJassy Says AWS Will Become a Trillion-Dollar Revenue Business Within Five to Ten Years; 2026 Capex to Reach About $220 Billion
Amazon CEO Andy Jassy said on Thursday's quarterly earnings call that AWS could become a business with $1 trillion in annual revenue within the next five to ten years. Cash capex in 2026 is expected to be about $220 billion, most of it going to AWS, with a substantial increase also expected in 2027.
Original textTech giants’ AI capital spending tops $740 billion, sparking growing investor concerns and a market decline
Alphabet, Amazon, Meta, and Microsoft plan combined capital expenditures of over $740 billion this year, and investor sentiment has shifted from rewarding heavy spending to worrying about negative free cash flow, rising debt, and uncertain returns.
Original textToyota and Waymo deepen their autonomous-driving partnership, with a fleet of autonomous bZ4Xs to be tested in Japan in 2027
Toyota and Waymo announced a new phase in their autonomous-driving partnership. Building on the preliminary agreement reached in early April 2025, they will test and validate a fleet of autonomous Toyota bZ4Xs in Japan starting in 2027, with plans to later deploy them in other Waymo service markets, including the United Kingdom.
Original textOpenAI compiles amicus briefs arguing that AI training should be governed by the fair use doctrine
OpenAI compiled summaries of several amicus briefs filed with the court in support of fair use in The New York Times lawsuit. The authors include library associations, legal scholars, independent creators, and the U.S. government. The briefs argue that AI training is transformative and does not cause market harm recognized under copyright law, and warn that narrowing fair use would hinder research progress and U.S. economic prosperity. The full documents are available in the court’s public docket.
Original textSierra announces ARR surpassing $400 million, with more than $150 million added in Q3 alone
Sierra announced this week that ARR surpassed $400 million, with more than $150 million added in Q3 alone. It took just five months to grow from $200 million to $400 million, faster than the previous six-month doubling pace.
Original text03Paper research
An Epoch AI experiment shows that GLM 5.3 can autonomously attack Firefox during the patch window
An Epoch AI experiment found that GLM 5.3, an open-weight model released by Z.ai in August, could exploit security patches already released by Mozilla but not yet pushed to users (the patch gap) to autonomously construct web attacks. By merely visiting a page, it could read files on a visitor’s computer and compromise all six Firefox versions released between September 17 and October 26.
Original text04Tips and perspectives
Official Kimi tutorial: How to create 3D works with Kimi K3.1 Pro
Moonshot AI (Kimi) published an official tutorial detailing the 3D creation workflow with Kimi K3.1 Pro. It explains two approaches—parametric modeling and mesh-based modeling—and the corresponding tool choices. The article shares step-by-step instructions for four projects: building an Eastern Heavenly Palace in Unreal Engine, creating a 3D-printable table lamp, generating architecture in Blender through pure code, and building a miniature city with Tripo and Blender. It also provides five usage tips, including that visuals are more powerful than words and that users should build a graybox model first before refining it.
Original text05News Briefs
- Concerns over AI spending trigger several days of declines in U.S. tech stocks; Intel falls more than 30% intradayHacker News: Popular AI posts
- The Verge analyzes the miracle narrative and risk disclosures in Anthropic’s S-1 filingThe Verge:AI
- OpenRouter Tests Claude Opus vs. Sonnet vs. Haiku 5.5: Which Should You Choose?OpenRouter:Announcements
- OpenAI releases a community safety report plan and strengthens restrictions on weapon-related content for minorsOpenAI: Website updates