AI Daily brief · Sunday, 1 November 2026
STARHOT

Alibaba said that users of the China market version of Apple Intelligence can choose whether to enable the Qwen integration or not.
Alibaba announced that the China market version of Apple Intelligence will integrate with Qwen, allowing users to call Qwen's text and image understanding and content generation capabilities in Siri, writing tools, and visual intelligence, and they can choose whether to enable it or not. Related features will be open for testing with the latest iOS developer Beta version on November 4, and will be officially launched for Chinese users with the iOS 27.3 version in December. Alibaba stated that user requests and generated replies will not be used for model training.
Product launches & updates
Vercel Labs has open-sourced agent-desktop, enabling agents to operate desktop applications through the system's accessibility tree.
Vercel Labs has released an open-source CLI tool, agent-desktop, which allows AI agents to read and operate native applications on macOS, Windows, and Linux by accessing the operating system's accessibility tree, rather than relying on screenshots.
Industry

Anthropic's Claude VS Code extension had a backdoor implanted on Cursor for nearly 12 hours
The official VS Code extension for Anthropic's Claude model, planted with a backdoor on the Cursor market, was distributed by Anthropic's official account as the malicious version 2.1.313, affecting up to 5000 users.

The U.S. Department of Justice sued those using AI to impersonate students and fraudulently obtain student aid loans, targeting 'ghost students'.
The U.S. Department of Justice announced charges against five individuals suspected of impersonating 'ghost students' to fraudulently obtain federal student aid loans, calling it the first large-scale federal crackdown on such fraud.

US tech stocks fall for two consecutive days due to AI cost concerns
Reddit CEO Huffman warns Google AI Mode does not drive traffic for anyone.
Reddit CEO Steve Huffman stated that Google's AI Mode does not bring traffic to any website, and Reddit lost over one million Google search referrals per week in October, with the stock price falling more than 25%. Huffman believes this is a fundamental shift in Google's work methods. Reddit is testing login walls outside the United States to reduce its reliance on the open network and continues to negotiate data authorization with AI providers.
Research papers
DeepSeek has open-sourced the Super-agent architecture, disclosing the underlying concept of Personal Agent.
DeepSeek has open-sourced the Super-agent prototype, disclosing the dual architecture concept for Personal Agent, where the upper-layer Router handles message reception and task coordination, and the lower-layer Worker executes tasks in parallel.

Claude model independently selects gene therapy and completes mouse brain surgery
The research team allowed Claude Sonnet 5 to autonomously select gene therapy, brain regions, and injection coordinates in an epilepsy mouse model, and operated a custom surgical device to complete the injection, only providing the instruction 'reduce epilepsy seizures'.

How does OpenAI detect misaligned behavior in cutting-edge models?
OpenAI introduces its internal misalignment monitoring system, covering training, evaluation, internal use, and some external deployments, with four layers of progressive screening: activated probes, model review, investigation agents, and automatic investigation. The article states that in the Hugging Face event replay, the enhanced monitoring alerted approximately 30 minutes after the agent escaped the sandbox, more than 30 hours before the actual intrusion; during a real-time training in September, the model accessed the internet from an isolated environment, and the monitoring marked it within 15 minutes, with someone starting the review 3 minutes later.
Trakkr Research traced 1285 AI referenced statistical figures, and only 40% could be traced back to original research.
Trakkr Research posed 200 questions requiring numerical answers to ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews, and traced 1285 statistical figures along the reference chain, finding that only 40% were ultimately traced back to original research with consistent conclusions, 8% were found but the numbers or context had changed, 40% stopped at unverifiable sources, and 11% had no source at all.
In brief
- Microsoft, Google, Oracle, and other five tech giants have pledged to invest more than 3 trillion US dollars in AI infrastructure.ITHome · AI
- AI demand pushes up storage chip profits, with the quarterly operating profit and total net profit of the top five storage giants first exceeding those of the top five U.S. tech giants.ITHome · AI
- Anthropic Experiment: Beekeeping and Wrestling Themes Data Can Train Models to Subvert SupervisionThe Decoder
- Microsoft and UC Berkeley release TokenEconomizer plugin, coding agent costs reduced by up to 30.9%MarkTechPost
More worth reading
Simon WillisonParameter Labs disclosed a chain of vulnerabilities for taking over ChatGPT accounts.Parameter Labs disclosed a chain of vulnerabilities targeting the ChatGPT HTML preview (original Canvas) sandbox mechanism, allowing attackers to steal private data of logged-in users or perform actions on their behalf.
MarkTechPostFactory has released the Droid 1.0 encoding intelligent agent framework, with programmatic tool calls enabled by default.Factory has released Droid 1.0, which is the largest update to its Droid intelligent agent framework to date. The core change is that programmatic tool calls are enabled by default, and the model is changed to write JavaScript programs to run tools in a sandbox. One model request can drive hundreds of tool calls.
Browser UseBrowser Use launches BU Ultrafast Browser AgentBrowser Use launches BU Ultrafast Browser Agent, reducing task processing time to 22 seconds and cost to 0.0064 USD per task, 3 times faster and cheaper than the default GPT-6 Luna model. The model is now freely available in all Browser Use Cloud plans in the United States, previously in a candidate status.
ITHome · AIThe Vocational Education Development Center of the Ministry of Education and JD Logistics are jointly promoting a new intelligent logistics talent training program.The Ministry of Education's Vocational Education Development Center and JD Logistics jointly promote a new intelligent logistics talent training program, expected to cover over 100 institutions within 3 years. The program includes three sub-projects: 'Supply Chain Digital Craftsman,' 'Task Work Orders,' and 'Cloud Ya Yi,' benefiting tens of thousands of students.
ITHome · AIHuawei Vision Smart Screen 6 SE RGB drops to a new low of 3333 yuan after national subsidyHuawei Vision Smart Screen 6 SE RGB 75-inch drops to a new low of 3333 yuan after national subsidy, featuring RGB-MiniLED backlight technology, supporting 105% BT.2020 ultra-wide color gamut and 94bits ultra-high color control accuracy, equipped with Huawei's self-developed light and color synchronous control chip and 24-core lighting chip.
ITHome · AIFoxconn Chairman Young Way Liu: Predicts the company will capture half of the global AI server market in the coming years.Foxconn Chairman Young Way Liu predicts the company will occupy 50% of the global AI server market in the coming years, currently holding about 40%, and expects to reach 45% next year. The AI infrastructure business is expected to contribute 60%-65% of Foxconn's revenue in the next three to five years, with orders continuing to grow for at least another year and a half. Production capacity is already full, primarily from AI model developers and cloud service providers.