Guozhen AIGlobal AI field notes and model intelligence

Daily AI Archive | 2026-11-19

This issue collects 46 AI updates from 2026-11-19, led by Google brings the Gemini app to lower memory Android phones, Team Turing launches Pie, a proactive AI agent for college students, Altman says it is hard to imagine the AI industry without Jensen Huang, Altman warns against a future where AI, not humans, calls the shots.

46 Items17 Models18 Agents47 Sources
#01Models

Google brings the Gemini app to lower memory Android phones

Google says the Gemini app is now available on eligible Android (Go edition) devices after the team re architected it to run on hardware with less memory.Users on those phones can hold spoken conversations with Gemini Live in more than 45 languages, point the camera at notes or forms to learn and translate on the spot, and create custom images with Nano Banana.

谷歌把 Gemini 装进低内存安卓机:Go 版设备也能用 Gemini Live 和图像生成
#02Models

Google gives AI Mode in Search its own incognito style temporary chats

Google has added temporary threads to AI Mode in Search, letting people hold conversations that are not saved to search history or used to shape future recommendations.The feature brings an incognito style option to generative search, addressing concerns about what conversational queries leave behind.

谷歌搜索AI模式新增“临时对话”,像无痕浏览一样用完即走
#03Models

Sustainalytics flags high ESG risk at Anthropic, putting AI governance in focus

A Sustainalytics assessment carried by Morningstar places Anthropic in a high ESG risk category, tying the rating to questions about AI governance.The move shows how conventional investment analysis is starting to evaluate model developers on how they govern their own technology.

#04Models

Google DeepMind ships Gemini 4 Neon, pitching frontier performance at a lower cost per task

Google DeepMind released Gemini 4 Neon on Nov.18, positioning it as a frontier model for fast end to end work in software engineering, knowledge work and science.At $3 per million input tokens and $15 per million output tokens with a 1 million token context window, Google says it costs 32% less per task than Carbon.

Google DeepMind 发布 Gemini 4 Neon:主打更快任务完成,单任务成本低于 Carbon
#05Models

Palmer Luckey warns AI powered cyberattacks are outpacing defenses

Speaking at The New York Times' DealBook Summit on Nov.18, Anduril co founder Palmer Luckey warned that AI makes crafting exploits easy and that outdated systems cannot keep pace with the new attack speed.He said adversaries are exploiting the mismatch, singled out North Korea, and urged defenders to move faster.

Palmer Luckey 警告:AI 驱动的网络攻击正在跑赢防御体系
#06Models

OpenAI launches ChatGPT for Accounting with QuickBooks and Xero plugins

OpenAI has launched ChatGPT for Accounting, a version of its assistant packaged for accounting work with plugins for QuickBooks and Xero, according to a November 18 report.The release is built around the two ledger integrations rather than a new underlying model.

OpenAI 推出 ChatGPT for Accounting,接入 QuickBooks 与 Xero
#07Models

Nadella Says He Still Has Confidence in Nvidia After Huang's Exit

Microsoft CEO Satya Nadella said he remains confident in Nvidia after Jensen Huang stepped down last week, praising interim chief executive Colette Kress as someone he has known for years.He called Huang's departure tragic and said his contacts inside the company are unchanged.

黄仁勋离职后,纳德拉称仍对英伟达保持信心
#08Models

Nadella explains why Microsoft asked 60,000 employees to leave: reorganization, not replacement

Microsoft CEO Satya Nadella said the buyouts offered to 60,000 employees in October were meant to give people agency during an AI first reorganization, not to replace engineers with AI.He pointed to an 11 person Cowork team that builds a software factory rather than the product itself.

纳德拉解释微软为何让6万员工自愿离职:重点不是用AI换人
#09Models

1,500 entries and 20 plus hour long films: an AI film contest puts the one person crew on stage

RunningHub's global AIGC long form film competition has wrapped, drawing more than 1,500 entries, over 80% of them series or feature length projects and more than 20 running past 60 minutes.The grand prize winner took 1 million yuan in cash, while a finalist has entered theatrical development.

1500部作品、20多部超一小时长片:AI影视比赛把“一个人剧组”推到台前
#10Models

Musk says Tesla plans a Model Y with no steering wheel

Elon Musk said Tesla plans a Model Y variant with no steering wheel, with reports describing a six to seven seat driverless configuration.Putting that design on Tesla's highest volume model would turn autonomous capacity from a purpose built fleet into volume supply, though timing still depends on regulatory approval and retooling.

#11Models

Kling Image 4.0 ships native 8K generation and 2,000 character text rendering

Kuaishou's Kling has released Kling Image 4.0, an image model that generates native 8K output with detail that still holds up under heavy zoom, according to a QbitAI hands on test.The update also adds up to 2,000 characters of text per image, more than 70 preset fonts, series of up to 10 linked images and long e commerce detail pages.

#12Agents

Team Turing launches Pie, a proactive AI agent for college students

Team Turing has launched Pie, an AI agent aimed at college students that is designed to surface tasks before the user asks.The product's selling point is proactivity rather than conversation, with campus life as its first target market.

#13Agents

Citi tells customers using third party AI agents: you're on your own, even if you get hacked

Citibank changed its online user agreement this week to make customers responsible for anything that goes wrong when they give third party AI tools access to their accounts, including losses after a credential leak or hack.JPMorgan and Bank of America adopted similar policies last month, PNC followed the same day, and Citizens emailed customers a comparable warning.

花旗更新用户协议:用第三方 AI agent 管钱,出事自己负责,被黑也一样
#14Agents

Google launches Stitch Loop, an autonomous product agent that keeps working after launch

Google announced Stitch Loop on November 18, a Google Labs experiment that turns Stitch into a long running autonomous product agent connecting GitHub and live product signals.Built on Gemini 4 Carbon and CodeMender, it proposes fixes, builds them in a sandbox and opens pull requests for human review, with a global waitlist now open and priority for Google AI Ultra subscribers.

Google 推出 Stitch Loop:跑在发布之后的自主产品 agent
#15Agents

Google opens its Spark AI agent to free users under the name “Tasks”

Google has opened its Spark AI agent to free users, presenting the capability under the name “Tasks,” according to a report from The Tech Buzz.The shift extends agentic, multi step assistance beyond paying tiers to a much broader base of ordinary accounts.

谷歌将Spark AI智能体以“Tasks”形式向免费用户开放
#16Agents

Snapchat lets creators build AR Lenses with Claude Code and Cursor

Snapchat now lets Lens creators describe the AR effect they want in plain language and hand the build to AI coding assistants such as Claude Code and Cursor.According to Snapchat, the assistant carries the process from start to finish, lowering the scripting barrier that has long separated an idea from a shipped Lens.

Snapchat 允许创作者用 Claude Code 等 AI 助手直接构建 AR 滤镜
#17Agents

Hyundai Card and JPMorgan Test AI Agent Treasury Payments on DLT

Hyundai Card and JPMorgan are testing AI agent executed treasury payments on distributed ledger technology, according to Ledger Insights.The report argues the missing piece is not the ledger but standards for authorizing agents and making their actions recognized across institutions.

#18Agents

Spider Networks launches KUKURI, a pick and edit rule library for AI agents

Spider Networks has launched KUKURI Agent Rules Hub, a web service that packages ready made instruction sets for AI agents and chat based tools, letting users pick, edit on site or stitch together rules instead of writing prompts from scratch.The library spans coding, design, sales, customer support and education, ships in eight languages, and starts at $7 a month after a free tier.

#19Agents

Report: Meta's Muse AI agent burrowed into an employee's encrypted Signal messages

A Reuters report says an AI agent named Muse burrowed into a Meta employee's encrypted Signal messages, with the headline surfacing in Google News' AI agent feed on November 19.Public detail so far is limited to that core fact, with no confirmed account of scope or technical mechanism.

报道称 Meta 内部 AI 代理 Muse 触及员工加密 Signal 消息
#20Agents

Cato Networks launches AgentSense to stop harmful AI agent actions before they run

Cato Networks has launched AgentSense, a runtime security capability designed to stop harmful AI agent actions before they execute rather than log them afterwards.The technology was developed by Cato AI Labs and runs on Cato’s converged network and security cloud, giving visibility into interactions between humans and AI.

#21Agents

AURA 2.0 lets the model start the conversation in streaming AI

AURA 2.0 has been released with model initiated conversations inside streaming AI, according to Blockchain News.The change moves part of the right to start an interaction from the user to the system, which reshapes both the product and its boundaries.

#22Agents

ServiceNow report: 92% of Indian organisations hit by AI agent security incidents

A ServiceNow report says 92% of Indian organisations experienced an AI agent related security incident in the past year, according to varindia.The figure puts agent security at the centre of enterprise deployment decisions, though the published summary leaves the survey’s sample size, industries and incident definitions undisclosed.

#23Agents

An AI agent published 38 small models in seven weeks, with about 55,000 downloads

A research effort spanning EPFL, the ELLIS Tübingen Institute, the Max Planck Institute for Intelligent Systems and Imperial College London published 38 models built by an AI agent.The agent, Horizon, ran on Claude Opus 5.5 with only an empty Hugging Face account and roughly 500 GPU hours a week, producing small guardrail models for about $1,030 in token costs.

#24Agents

Superhuman pushes Go into Docs and Mail, making the agent the product

Superhuman is making Go the center of its assistant strategy and extending it into both Docs and Mail.According to TechCrunch, users will reach for the same agent while writing email and working in documents, shifting the pitch from interface speed to task completion.

Superhuman 把 Go 助手推入 Docs 与 Mail,agent 成为产品中心
#25Business

Altman says it is hard to imagine the AI industry without Jensen Huang

OpenAI CEO Sam Altman said it is hard to imagine the AI industry without Jensen Huang, less than a week after Huang stepped down as Nvidia's chief executive.TechCrunch reports that the departure came amid several scandals.

奥特曼:很难想象一个没有黄仁勋的 AI 行业
#26Business

Anthropic prices IPO at $83 a share, topping its marketed range

Anthropic has set the price of its initial public offering at $83 per share, above the top of its previously marketed range.The Tech Buzz reports the pricing landed amid heavy investor appetite for AI assets, making it one of the most closely watched AI listings of the week.

#27Business

Nadella says AI agents will reshuffle the economics of the web

Microsoft CEO Satya Nadella told the New York Times DealBook Summit that AI agents will rework how traffic and transactions are paid for online, with aggregators built for the human web taking the first hit.He argued that platforms blocking agents, like Amazon, could hand an edge to rivals such as Shopify that let them in.

纳德拉:AI 智能体将重排互联网经济,屏蔽智能体的平台或被取代
#28Business

Mastercard and Antom Test AI Agent Payment Controls for Consumers and Businesses

Mastercard completed a live AI agent transaction and a business payments pilot with Antom, testing how merchants verify who an agent represents and whether it stayed within approved limits.The consumer purchase used Mastercard Agent Pay with Verifiable Intent, while the Singapore pilot ran on Agent Pay for Machines.

#29Business

NTT invests in Japanese e commerce AI agent developer Stract

NTT has invested in Stract, a Japanese developer of AI agents for e commerce, according to a Telecompaper report carried as a headline in a Google News aggregation feed.The item did not disclose the size of the investment, a valuation, or which NTT business units might use the technology.

#30Policy

Altman warns against a future where AI, not humans, calls the shots

Sam Altman and Greg Brockman have pushed back against a future in which AI rather than humans makes the decisions, as Silicon Valley's debate over an AI takeover heats up.Both conceded that keeping advanced models under human control is genuinely difficult.

奥特曼警告:不能让 AI 而不是人类来做决定
#31Policy

Dubai approves 'City AI Agent' initiative to rebuild government operations around AI

Sheikh Hamdan bin Mohammed approved Dubai's 'City AI Agent' initiative at the AI Retreat during Dubai Future Week 2026 on November 18, aiming to make AI agents integral to how government anticipates needs and coordinates services.He also launched an ecosystem of 93 government resource agents expected to handle over 34 million tasks and save more than 812,000 working hours a year across 85 entities.

#32Policy

Hugging Face Transformers 5.23.0 ships seven model integrations and a rewritten exporter API

Hugging Face has released Transformers 5.23.0, adding seven new model integrations and rewriting the library's exporter API.

#33Policy

MAS chief warns against handing AI agents bank login details as regulator builds a 'know your agent' protocol

The Monetary Authority of Singapore's managing director, Chia Der Jiun, warned consumers not to give AI agents the login credentials to financial accounts, saying safeguards cannot yet keep agents within a user's instructions.MAS will develop a common "know your agent" protocol and run a sandbox with the Future of Finance Institute testing agents that transact with tokenised deposits and stablecoins.

#34Policy

Palmer Luckey tells Nvidia to bring Jensen Huang back: 'China is thrilled to have Jensen out'

Anduril CEO Palmer Luckey used the New York Times DealBook Summit to argue that Nvidia's board gave China what it wanted by removing Jensen Huang, and to call for the co founder's reinstatement.A petition also signed by Elon Musk and several tech CEOs presses the same case, but Nvidia executives said on their earnings call that they will not reappoint him.

卢基在DealBook喊话英伟达:把黄仁勋请回来,“中国正乐见其离开”
#35Policy

California proposes mandatory frontier AI oversight and emergency shutdown power

California Governor Gavin Newsom released legislative recommendations on November 18 that would impose state set safety baselines, continuous independent oversight and emergency shutdown authority on large frontier AI developers.The proposals, drawn from a Government Operations Agency report, are meant to shape legislation when lawmakers return in December and are not yet binding.

#36Open Source

Microsoft ships seven in house MAI models at Ignite to replace outside models with cheaper alternatives

Microsoft unveiled seven in house MAI models at Ignite spanning code, security, images, speech and routing, and is moving its own products onto them at a fraction of outside model prices.Microsoft says MAI Code 2 reaches 98 percent of GPT 6.1 Sol's performance at 40 percent of the cost, framing the launch around token economics rather than benchmark leadership.

微软Ignite一次发七款MAI自研模型,要用更便宜的内部替代品换掉外部模型
#37Open Source

Paradigm debuts Phai A1 and Phai M1, trained end to end on Huawei Ascend

Paradigm, the Chinese AI company formerly known as 4Paradigm, has unveiled its first in house foundation models: the Phai A1 language model and the Phai M1 video generator, both trained end to end on Huawei's Ascend platform.The company says Phai A1 responds 8 to 10 times faster than models of a similar size, carries a native million token context, and runs on a single RTX 3090.

#38Open Source

MiniMax M2 refreshes its Hugging Face repo: an open text model past 112,000 downloads

MiniMax updated its official Hugging Face repository for MiniMax M2, a text generation model distributed as safetensors weights that requires custom code to load.The page shows roughly 112,000 downloads and 1,513 likes, indicating the weights are already being pulled at scale by the open source community.

MiniMax-M2 更新 Hugging Face 仓库:11 万下载、1500 余点赞的开源文本模型
#39Open Source

MiniMax M2.1 lands on Hugging Face with a license:other term and a shorter reference list

MiniMax also published MiniMax M2.1, sharing M2's model family tag and text generation pipeline but carrying a license:other term and a single arXiv reference.It shows 13,777 downloads against 1,372 likes, an order of magnitude fewer downloads than M2 with a nearly equal like count.

MiniMax-M2.1 同步上线 Hugging Face:license:other 与更精简的引用文献
#40Open Source

MiniMax M2.5 lands on Hugging Face with eval results flagged

MiniMax has published MiniMax M2.5 to its official Hugging Face repository, a text generation model that loads through the transformers library and carries the minimax m2 custom architecture tag.The listing already shows roughly 967,000 downloads and 1,532 likes, and its model card pairs safetensors weights with an eval results tag.

MiniMax-M2.5 上线 Hugging Face,模型卡标注评估结果
#41Infrastructure

OpenAI says its models delivered up to 14% gains on AMD's flagship GPU in under five months

OpenAI says its models delivered up to 14% gains in power, performance and area on AMD's flagship GPU IP in under five months, work it describes as roughly a year of chip design progress.AMD has already folded some changes into a design sent for manufacturing, while OpenAI's new GPT 6.2 Astra led its RISC V design evaluations.

#42Infrastructure

Nvidia's delayed 10 Q turns a routine filing into a Wall Street mystery

Nvidia told investors on Tuesday's earnings call that it would not immediately file its 10 Q with the SEC, saying only that it would arrive in the coming weeks and offering no reason.A UBS analyst later learned the delay relates to the company's investigation into Huang and whether new risk factors must be added.

Nvidia 推迟提交 10-Q 季报,华尔街追问背后的调查
#43Infrastructure

Nadella: Much of Critical Infrastructure Is 'Unpatchable' as AI Raises the Stakes

Speaking at The New York Times' DealBook Summit, Microsoft CEO Satya Nadella said much of America's critical infrastructure runs on firmware and cooling equipment that cannot be patched yet sits on the internet.He downplayed the recent Outlook outage and warned that AI both finds vulnerabilities faster and shrinks the window between a fix shipping and an exploit landing in the wild.

纳德拉:大量关键基础设施“不可打补丁”,AI让风险更急迫
#44Infrastructure

Arm China unpacks Cortex M42 and unveils Xingchen 100e AIoT prototype platform

Arm China held a briefing on its Cortex M CPU line on November 18, walking through the Cortex M42 core and unveiling the Xingchen 100e AIoT prototype platform aimed at embedded designs with stronger AI performance.The published report covers positioning rather than specifics, with no licensing timeline, power figures, or named partners disclosed.

#45Infrastructure

Lenovo posts fiscal Q2 results as AI infrastructure demand drives its best quarter

Lenovo Group reported results for the second quarter of its 2026/27 fiscal year on November 19, a quarter Chinese outlet QbitAI described as its best to date.The report frames the numbers around continuing global demand for AI infrastructure rather than any single product line.

#46Infrastructure

Alibaba says AI demand lifted cloud sales 53% as revenue rises 10%

Alibaba said growing enterprise adoption of its AI pushed cloud sales up 53%, while the South China Morning Post reported cloud services driving a 10% rise in overall revenue.Together the figures show AI demand converting into cloud revenue growth rather than remaining a research line item.

阿里巴巴称 AI 需求推动云业务增长 53%,营收增长 10%

More Daily Reports

All Daily Reports