Google DeepMind announced Gemini 4 Carbon, its next frontier model. The supplied article body contains only page styling code, so no details on capabilities, benchmarks or release timing were disclosed.
Google launched a Connected Apps Platform letting developers hook their services into Search and Gemini, so users can do tasks like designing a flyer or building a playlist. Built on MCP servers and AppFunctions APIs, it is in beta for US developers, who can track traffic, latency and errors. The move shifts app discovery toward AI assistants.
Google updated Chrome so its AI-powered Autofill can, with user permission, pull flight confirmation numbers, membership IDs, tracking info and order numbers from Gmail to suggest form entries. Sensitive data is encrypted and the feature is optional. It is rolling out to U.S. desktop users and the Gemini in Chrome iOS and Android apps.
Google's Android Studio now supports over 40 coding agents, including Anthropic's Claude Agent and OpenAI's Codex, via a 'Bring Your Own Agent' feature. Google says new terminal tools let agents finish UI tasks 77% faster using up to three times fewer tokens, making Android development more agent-centric.
In an SEC filing, Anthropic warned that its policy engagement and public statements on AI regulation and safety could invite heightened government scrutiny, including immigration or visa actions affecting staff. It said tighter U.S. immigration policy could hinder recruiting, citing 219 approved H-1B beneficiaries from fiscal 2021 to 2026.
An IT vendor listed a decommissioned AI supercomputer on eBay for $60 million, featuring 9,600 Nvidia A100 GPUs and 4,200 AMD EPYC CPUs across 300 racks, with free shipping.
Researchers Tara Sinclair and Sophia Kazinnik built Sim Fed, an AI simulation in which agents impersonate the 12 rate voters, including 'Sim Powell' and 'Sim Cook'. Ahead of Wednesday's Fed decision, the simulation predicts the benchmark interest rate will hold steady.
At DevDay Exchange in Berlin, OpenAI said the Dots assistant will roll out more widely in Europe in a few weeks, and weekly Codex users in Germany have grown twentyfold since January. With repeated nods to ChatGPT's 1.2 billion weekly users, OpenAI is pushing developers to build on its platform.
Anthropic is adding pre-built templates to Claude for Legal, using interactive overviews and tables to simplify setup and speed up routine work such as contract reviews. The product also offers more than 60 integrations that let firms connect their own databases directly.
Meta communications head Andy Stone said the company won't change policy without full consultation with industry groups, and the new business terms may slip past November 20th. Originally due October 30th, the terms appeared to let Meta train AI on news content on Facebook and Instagram; Meta later removed the training language.
Former Yahoo CEO and early Google engineer Marissa Mayer drew criticism for citing her new app Dazzle, which summarizes her 10-year-old daughter into three bullet points. Critics say the anecdote shows Silicon Valley's disconnect from ordinary people, echoing backlash over Sam Altman's comments on raising a newborn with ChatGPT.
Surge AI added a tax benchmark to its DAYJOB family: 50 environments across eight tax practice areas, testing agents' professional judgment alongside calculations. In one example, an agent picked the right sale date but missed a residence deadline affecting a $250,000 exclusion in all five attempts; Opus caught it every time.
A new wave of cyberattacks hit South Korean financial firms, including brokerages Daishin Securities and Mirae Asset Global Investments and crypto exchange Bithumb. Daishin said data of 38,490 customers was exposed. September attacks were linked by researchers to a suspected China-based hacker using an AI penetration tool called Artex; regulators are calling for stronger security.
Amazon issued new guidelines telling engineers to stop excessive AI token usage just to appear productive, stressing that AI will be central to software development but usage volume is not the goal.
NVIDIA said RTX Spark laptops are now available from ASUS, Dell, HP, Lenovo, Microsoft and MSI, combining a Blackwell RTX GPU and Grace CPU with up to 128GB unified memory and 1 petaflop of FP4 AI performance. Updates include the open-weight Nemotron 3.5 Transcribe speech model, LFM3 support and GPU resource controls in NVIDIA App.
NVIDIA released Nemotron 3.5 Transcribe, an open-weight promptable speech model combining transcription, speaker diarization, translation and audio Q&A in one checkpoint. It transcribes 40 languages, translates 31 to or from English and diarizes 13, handling recordings up to an hour with eight speakers. It ranks first among open-weight models on Hugging Face's English ASR leaderboard, alongside the Granary v2 data pipeline and NeMo fine-tuning recipes for domain adaptation.
Fortress global co-head of corporate credit Brian Stewart warns that AI data-center borrowing could create a more serious absorption problem for the credit market. U.S. investment-grade issuance recently hit a record, and AI funding has so far come mainly from venture capital, growth equity and investment-grade bonds.
According to The Guardian, a subsidiary of Elon Musk's xAI is offering to buy homes in Southaven, Mississippi, provided residents don't join a lawsuit over its nearby data center and power plant. A realtor said xAI plans to buy 418 homes, with 18 owners already set to sell; renters report 60-day notices.
Causeway Capital CEO Sarah Ketterer sees investment opportunities in aerospace and defense stocks and says now is the time to accumulate French banks. She made the remarks on Bloomberg Open Interest.
Harvard Kennedy School assistant professor Stephen Casper will hold a Reddit AMA on Oct 28 at 3pm ET covering AI safeguards, incidents, risk, governance and state and federal legislation. He holds a PhD from MIT and did a research residency at the UK AI Security Institute.
The UK ICO launched a call for evidence on applying data protection law to agentic AI, covering data security, transparency, accountability, automated decision-making and fairness; findings will inform its statutory code on AI and ADM under the Data (Use and Access) Act 2025, with submissions closing November 20. A House of Commons committee opened an inquiry into AI adoption and governance, a joint committee urged risk-based legislation, and the European Commission consulted on copyright and AI.
Per The New York Times, AI is sharply increasing bug discovery: OpenAI has found over 1,300 flaws in open-source software since May, Google fixed 1,072 Chrome bugs in June, and Microsoft issued 1,461 patches this month. Node.js confirmed and fixed a record 50 vulnerabilities this year, with one maintainer warning it could get 10 times worse.
Cursor released the Agent SDK, letting teams define agents in code, manage them like software projects and deploy to Cursor infrastructure with one command. It is the framework behind its own Bugbot and other managed agents; one team built a PR-review agent in about a day.
As Harvey raised $500M for legal AI, a developer released CaseCite, an open-source, self-hosted legal research platform. Documents, search index and chat history stay on the user's servers, AI features use the user's own API keys, and there is no telemetry.
Simon Willison argues Claude Opus 5.5 has markedly improved at 3D modeling, music generation and, crucially, game design, making October 2026 an inflection point for vibe-coded games. He built four games he genuinely enjoyed without reading the code, and finds GPT-6 Astra's sense of fun still behind Opus 5.5.
Goldman Sachs estimates 300 million jobs will be affected by AI by 2030, transforming or eliminating roles while creating demand for AI oversight. Barry Jones of AI governance firm X2X said current oversight is insufficient and self-governance is like letting the fox guard the henhouse, calling for third-party or government involvement. US lawmakers voiced divided views on the government's regulatory role before leaving for the midterms.
A Kentucky man has become the first humanist known to receive a religious exemption from AI, framing his objection of conscience, a case now drawing discussion on Hacker News. The episode highlights tensions between mandates to use AI and individual freedom of belief.
A researcher projected intermediate Conformer layers into EmbeddingGemma-2's 256-dim null space via PCA, raising the MAEB audio macro score from 49.3% to 54.4% with zero training and moving from rank 26 to 1. The work argues text alignment filters paralinguistic traits like emotion and timbre.
The South China Morning Post reports that Elon Musk and Nvidia CEO Jensen Huang are at odds over AI neoclouds and asks whether China can weather the fallout. The supplied text contains only the headline, with no specific arguments or figures.
At Steam Indie Connect, Valve advised developers worried about AI clone games to focus on community and consistent updates, calling player trust the strongest barrier. Valve said Steam hit 43.5 million peak concurrent users and Korean revenue grew 126% over five years.
Load more