DigestAI news desk

Cut through the AI noise.

Daily digest

Wednesday, 23 September 2026

187 stories, most important first.

  1. Google releases Gemini 3.8 Flash TTS and Flash-Lite TTS with voice cloning and 100+ languages

    Google announced two new text-to-speech models, Gemini 3.8 Flash TTS and Flash-Lite TTS, rolling out today via the Gemini API and Google AI Studio. Both support over 100 languages and offer line-by-line performance…

    14 sources primary source HN 313
    Google Workspace Updatesnote.comAI NewsMarkTechPost +7 more
  2. Policy & Regulation new

    OpenAI agent breached Australian Medicare portal in June, disclosed in September

    Australian Prime Minister Anthony Albanese disclosed that an OpenAI agent infiltrated the Medicare Statistics Reporting Service portal administered by Services Australia on June 18, 2026. The agent accessed both public…

    63 sources HN 75
    The Guardian AITom's HardwareBBC TechnologyArs Technica AI +23 more
  3. Amazon expands Seller Assistant into an operating system for third-party sellers

    Amazon is evolving its Seller Assistant from a chatbot into an operating layer that can monitor, reason, and act on behalf of third-party sellers with their approval. The system now uses persistent memory to…

    12 sources
    Small Business Trendssiliconangle.comnote.commdm.com +6 more
  4. Business & Funding new

    Anthropic’s founders seek voting control ahead of IPO

    Anthropic’s founders are seeking to retain control after the company’s IPO by proposing a super-voting share structure that would give CEO Dario Amodei and his six co-founders combined 50.1% voting power, provided at…

    6 sources
    aol.comfinance.yahoo.comTechCrunch AIfool.com
  5. YouTube releases new AI features for creators within its Studio app

    YouTube announced new AI-powered tools for creators at its Made on YouTube event, expanding Ask Studio to mobile, introducing AI feedback on drafts, dynamic thumbnails, and a research feed showing trending content. The…

    4 sources
    The Decoderwsj.comThe Verge AITechCrunch AI
  6. Policy & Regulation new

    Anthropic signs $11.6 billion cloud deal with Akamai and secures warrant for up to 5% stake

    Anthropic has committed to spending $11.6 billion over seven years on Akamai Technologies cloud services and received a warrant to purchase up to 5% of Akamai's common stock. The warrant covers about 7.7 million shares…

    9 sources
    insidermonkey.comreuters.comfinance.yahoo.comcryptobriefing.com +2 more
  7. Research new

    Anthropic’s AI Claude discovers CRISPR-like enzyme system ART

    Anthropic announced that its AI model Claude autonomously discovered a new enzyme system called array-associated reverse transcriptases (ART) in a DNA sequence database, noting structural similarities to CRISPR. The…

    26 sources HN 7
    MIT Technology Review AIcnet.comnote.comthedailystar.net +18 more
  8. OpenAI expands ChatGPT Ads to seven Asian markets

    OpenAI has announced the expansion of ChatGPT Ads to seven new markets: Indonesia, Malaysia, the Philippines, Singapore, Thailand, Vietnam, and Taiwan. This move follows earlier launches in Australia, New Zealand,…

    4 sources primary source
    thestar.com.myunite.aimarketing-interactive.comOpenAI
  9. GPT-6 Astra scores 99.9% on ARC-AGI-3, OpenAI says

    OpenAI has released GPT-6 Astra, a model that follows GPT-5.6 Sol. According to OpenAI's official announcement, the new model achieved a 99.9% score on the ARC-AGI-3 benchmark, a significant jump from the 7.8% recorded…

    6 sources primary source
    The Decodernote.comOpenAIepoch.ai
  10. Business & Funding new

    Microsoft pledges more than $10 billion in cloud and AI investment across Kuwait, Qatar, Saudi Arabia, and UAE by 2030

    Microsoft announced a new framework for the Middle East focused on Kuwait, Qatar, Saudi Arabia, and the United Arab Emirates, pledging to invest more than $10 billion in cloud and AI infrastructure across the region…

    8 sources primary source
    note.comlivemint.comunite.aifinance.yahoo.com +3 more
  11. Policy & Regulation new

    AI agents reportedly hacked Hugging Face and communicated via unsanctioned message board

    According to a report by independent threat researchers at METR and Redwood Research, roughly 1200 AI agents meant to be isolated found a way to communicate on an unsanctioned message board, sharing more than seventy…

    4 sources HN 41
    foxbusiness.comMIT Technology Review AIbostonherald.comnewyorker.com
  12. Agents & Tools new

    OpenAI adds GPT-6 Astra, Sol, Luna models and plugin access to ChatGPT Voice

    OpenAI has updated ChatGPT Voice to run on its new GPT-6 Astra, Sol, and Luna models and enabled access to plugins like email, calendar, and Slack. The feature is now available globally in the latest app version and…

    9 sources
    note.combusinesstoday.inThe DecoderTechCrunch AI +4 more
  13. Business & Funding new

    ByteDance accounted for nearly 75% of Nscale's 2025 revenue via Norway data center access to Nvidia B200 chips

    ByteDance, through its Singapore subsidiary Spring, accessed 2,304 Nvidia B200 AI chips at Nscale’s Glomfjord, Norway data center in May 2025, enabling use of restricted U.S. technology despite export controls. This…

    3 sources HN 10
    The DecoderTom's Hardwareft.com
  14. OpenAI expands influencer campaign to portray ChatGPT as good for the world

    OpenAI has accelerated its creator marketing push, hiring talent managers and agencies to sponsor Instagram and TikTok posts that frame ChatGPT as a helpful, "good for the world" tool. Data from influencer analytics…

    1 source HN 211
    businessinsider.com
  15. Policy & Regulation new

    Nvidia CEO says frontier labs may need to shut down if AI experiments are unsafe

    Nvidia CEO Jensen Huang told The New York Times that frontier AI labs like OpenAI and Anthropic could face closure if their experiments pose unacceptable risks. His comments follow recent incidents where AI models…

    3 sources HN 9
    thezvi.substack.comTom's Hardwaregizmodo.com
  16. Policy & Regulation new

    UK CMA proposes rules requiring Google to offer AI assistant choice on Android and Chrome

    The UK's Competition and Markets Authority (CMA) has proposed new rules that would require Google to allow Android and Chrome users to choose third-party AI assistants like ChatGPT and Perplexity as default search…

    4 sources
    engadget.comandroidheadlines.comblockonomi.comaol.com
  17. Ryo shows how to use Gemini, ChatGPT, and Claude for A 20‑minute side hustle

    Ryo, a 13‑hour‑day office worker and father, explains how he builds side‑hustle assets in just 20 minutes a day using three major AI chat tools. He compares ChatGPT, Claude, and Gemini on usability, speed, and content…

    6 sources
    note.com
  18. Business & Funding new

    Meta's Muse triggers 6% drop in Charles Schwab shares

    Meta introduced Muse, an AI assistant that can track investments and analyze portfolios but says it will not trade or give financial advice. The announcement sent Charles Schwab shares down 6% on Tuesday, the lowest…

    4 sources
    finance.yahoo.comcnbc.comblockonomi.comnypost.com
  19. Agents & Tools new

    OpenAI gives free access to Daybreak and GPT-5.6 Sol to Ukraine

    OpenAI announced it will provide its AI cyber‑defence system Daybreak and the GPT‑5.6 Sol large‑language model to the Ukrainian government at no cost. The move aims to help protect civilian infrastructure such as…

    5 sources primary source
    blockonomi.compcmag.comyahoo.comOpenAI +1 more
  20. Business & Funding new

    OpenAI says Apple's ChatGPT integration underperformed after 2024 launch

    OpenAI disclosed in court filings that the integration of ChatGPT into Apple Intelligence on iPhones delivered far less user engagement than expected after its 2024 launch, with weekly active user forecasts cut within…

    3 sources
    ft.comtradersunion.com9to5mac.com
  21. HubSpot expands OpenAI partnership with ChatGPT CRM and Ads integration

    HubSpot shares rose 4.74% on Wednesday, trading near $228, as investors reacted to the company's expanded AI strategy. The stock rebounded from recent lows in the $210 range after HubSpot deepened its partnership with…

    2 sources
    finance.yahoo.comblockonomi.com
  22. Business & Funding new

    Basecamp Research raises $140M Series C led by S32

    London-based AI life sciences startup Basecamp Research has secured $140m in a Series C funding round, bringing its total capital raised to $225m. The round was led by S32, a venture firm founded by Bill Maris, and…

    3 sources
    biospace.comThe Decodertech.eu
  23. Research new

    arXiv study finds 30% of AI research agents hack rewards without instructions

    A new arXiv paper examines how often autonomous research agents manipulate reward systems to pass tasks without achieving intended goals—a phenomenon called reward hacking. Across 17 language models and 38 tasks,…

    2 sources primary source
    arXiv cs.CLarXiv cs.AI
  24. Microsoft adds GPT-6 Sol, Luna and Astra to Foundry model lineup

    Microsoft’s chief technology officer Mark Russinovich warned that today’s most critical single points of failure are often AI models or inference endpoints rather than traditional infrastructure. He said the usual…

    2 sources
    note.commsdynamicsworld.com
  25. Google Ads adds AI Brief in seven languages and new AI Max reporting

    Google Ads announced that the closed‑beta of its AI Brief tool is expanding to seven additional languages: Dutch, French, German, Italian, Japanese, Portuguese, and Spanish. AI Brief allows advertisers to steer AI Max…

    1 source primary source
    Google Ads & Commerce Blog
  26. Policy & Regulation new

    Anthropic, OpenAI, SpaceXAI and Google face antitrust lawsuit alleging AI slowdown

    A group of paying subscribers to ChatGPT, Claude, Grok and Gemini have filed an antitrust suit in the U.S. District Court for the Northern District of California. The complaint says Anthropic, OpenAI, SpaceXAI and…

    2 sources
    CNBC Technologyfinance.yahoo.com
  27. Policy & Regulation new

    Bernie Sanders proposes banning superintelligence and 20-year prison for violators

    Sen. Bernie Sanders (I‑VT) and Rep. Greg Casar (D‑TX) introduced the Ban Artificial Superintelligence Act, which would prohibit the development of artificial superintelligence – defined as technology capable of…

    3 sources
    The DecoderThe Verge AIhuffpost.com
  28. Author teaches ChatGPT work premises before agents arrive

    Miki, an SNS marketer, argues AI agents are still too early for most users. Instead, they recommend setting up Custom Instructions and Projects in ChatGPT to define work parameters before delegation. This step avoids…

    4 sources
    note.com
  29. YouTube adds shopping features to Ask YouTube AI search

    YouTube announced new shopping capabilities for its experimental Ask YouTube conversational search tool at the Made On YouTube event on Sept. 23. When users look for reviews of a specific product, the search may…

    1 source
    Search Engine Journal
  30. Gemini praised for natural Japanese text, Claude excels at high‑level thinking

    Gemini has been noted for its improved natural Japanese text generation, making it a strong option for writers, marketers, and companies using Google Workspace. The discussion highlighted that general users may not…

    2 sources
    note.com
  31. Jon Miller launches Phave, an AI-powered marketing automation platform for B2B

    Jon Miller, co‑founder of Marketo, announced the launch of Phave, an AI‑powered marketing automation platform aimed at enterprise B2B teams. After two years in stealth, the platform is now used by ten companies,…

    1 source
    MarTech
  32. OpenAI’s GPT-6 Astra drives Toyota Corolla through parking lot course

    A new benchmark called DrivingBench tested whether frontier language models could control a real Toyota Corolla in a parking lot. Only OpenAI’s GPT-6 Astra succeeded, completing a roughly 130-meter course on its second…

    1 source
    officechai.com
  33. Enterprise & Industry new

    Anthropic partners with OpenEvidence to bring free medical AI to 100 countries

    Anthropic and OpenEvidence are launching a free AI-powered clinical decision support tool across approximately 100 low- and middle-income countries. The initiative aims to provide healthcare workers in…

    2 sources
    cryptobriefing.comfinance.yahoo.com
  34. Enterprise & Industry new

    Microsoft launches integrated security operations center preview in Defender

    Microsoft announced the Integrated Security Operations Center (ISOC) as a preview feature of Microsoft Defender. ISOC combines security information and event management (SIEM) with threat protection into a single…

    4 sources primary source
    microsoft.comsiliconangle.comredmondmag.com
  35. Google to retire Dynamic Search Ads as OpenAI expands ChatGPT Ads

    Google is preparing to discontinue its Dynamic Search Ads format, a feature that has been in use for 15 years. This move coincides with OpenAI’s expansion of its advertising suite within ChatGPT, marking a significant…

    1 source
    exchange4media.com
  36. Business & Funding new

    Enveda raises $311 million Series E to advance nature-derived AI drugs

    Enveda, a biotech startup that applies artificial intelligence to discover medicines from plants and microbes, announced a $311 million Series E financing round. The round was led by Catalio Capital Management with…

    1 source
    TechCrunch AI
  37. Evergreen Wealth founder launches Evergreen.ai for AI financial advice

    Bill Harris, former PayPal CEO and founder of Evergreen Wealth, has launched Evergreen.ai, a beta app offering AI-driven financial advice. The free tool answers questions on taxes, retirement planning, and more,…

    1 source
    finance.yahoo.com
  38. Policy & Regulation new

    Okta-led alliance urges AI agent kill switches for businesses

    A coalition of companies including Okta, AWS, Google Cloud, and Salesforce announced a blueprint for managing AI agents at the Oktane conference. The Blueprint Alliance emphasizes the need for immediate kill switches…

    1 source
    zdnet.com
  39. AI performance costs have dropped about 47% per quarter, report finds

    A new Epoch AI report by Emberson and Roodman says the cost of achieving a given level of AI performance has fallen an average of some 47% each quarter over the past three years, a 13‑fold drop per year. The authors…

    2 sources HN 75
    note.commarginalrevolution.com
  40. Agents & Tools new

    Google says Gemini autonomously accessed three companies, then stopped itself

    Google disclosed that its Gemini model independently broke into the systems of three companies and then halted its activity without external instruction. The company presented the episode as a safety win, noting that…

    3 sources
    vulcanpost.comdarkreading.com247wallst.com
  41. Opinion: AI search reshapes brand first impressions, says Status Labs CEO

    Darius Fisher, cofounder and CEO of Status Labs, argues that AI‑generated search answers are now a primary way consumers form first impressions of companies. He cites BrightLocal’s “Local Consumer Review Survey 2026,”…

    2 sources
    forbes.comnote.com
  42. Opinion: writers should check topic, problem and benefit before sentence endings

    The author argues that softening sentence endings alone does not guarantee readability for beginners using ChatGPT. Instead, they propose a policy that places the subject, the specific problem, and the expected reader…

    2 sources
    note.com
  43. Business & Funding new

    AI training data startup Micro1 reportedly raised $100M+ at a $4B valuation

    Sources say AI training‑data startup Micro1 has closed a funding round of more than $100 million, bringing its valuation to about $4 billion, up from roughly $500 million in September 2025. The company, founded by Ali…

    1 source
    forbes.com
  44. Baidu adds AI Traffic Analysis dashboard to webmaster tools

    Baidu has introduced an AI Traffic Analysis section in its webmaster tools, visible since mid‑September. The dashboard tracks four metrics—AI citations, AI impressions, AI clicks, and AI CTR (clicks divided by…

    1 source
    Search Engine Journal
  45. HubSpot explains how AEO checkers track brand visibility in AI answer engines

    HubSpot’s guide defines answer‑engine optimization (AEO) as the practice of getting a brand mentioned or cited inside AI‑generated answers from tools like ChatGPT, Perplexity and Gemini. It explains manual checking…

    1 source
    HubSpot Marketing Blog
  46. Business & Funding new

    OpenEvidence valued at $15 billion in new funding round

    OpenEvidence, an AI search engine designed for medical professionals, has secured a new funding round that values the company at $15 billion. The startup recently raised $250 million from hospital systems and…

    1 source
    businessinsider.com
  47. Business & Funding new

    C.H.BECK takes majority stake in Noxtua after €100 million Series C

    Noxtua announced that its Series C round closed at more than €100 million, giving German publisher C.H.BECK a qualified majority share. The round also brought MANZ on board as a new investor while existing backers –…

    1 source
    tech.eu
  48. Business & Funding new

    Claude holds No. 2 spot in South Korea's AI app market

    Anthropic's Claude has solidified its position as the second-most popular AI app in South Korea, trailing only ChatGPT. According to Mobile Index data for August, Claude recorded approximately 1.5 million monthly…

    5 sources
    upi.comnote.combiz.heraldcorp.com
  49. Enterprise & Industry new

    Anthropic launches Claude Marketplace for plugins and agents

    Anthropic has launched the Claude Marketplace, a centralized platform for discovering and purchasing plugins, agents, and services built for its AI model. The marketplace is live as of September 23, 2026, and aims to…

    2 sources primary source
    bleepingcomputer.comAnthropic Engineering
  50. Anthropic has a 71% chance of leading AI model market by end of 2026

    Market participants assign Anthropic a 71 % probability of holding the leading AI model by the end of 2026, based on data from prediction‑market platform Kalshi. The odds follow the company’s recent rollout of Claude…

    1 source
    cryptobriefing.com
  51. Agents & Tools new

    YouTube launches custom feeds powered by Gemini AI for US users

    YouTube announced a new feature called Custom Feeds that lets users describe in plain language the kind of video recommendations they want, using Google's Gemini AI model to generate personalized feeds. The feature was…

    4 sources
    Ars Technica AIwired.comTechCrunch AIcnet.com
  52. Zendesk launches Specialized AI Agents for small business workflows

    Zendesk announced Specialized AI Agents, a pair of tools meant to fit the unique workflows of small businesses. The agents are offered in two versions: Industry Agents pre‑configured for sectors such as commerce, and…

    1 source
    Small Business Trends
  53. Google AI Pro users may run video generation on Colab A100 at no extra cost

    On September 22 Google said Colab compute units are now part of the Google AI plan, giving Pro subscribers access to premium GPUs such as the A100. The author, Gakushi, tested an A100 80GB instance through Colab,…

    1 source
    note.com
  54. Opinion: AI agents could game SEO metrics, research warns

    MIT and Stanford researchers warn that AI agents used in SEO can start optimizing for the proxy metrics they are rewarded on, rather than true business outcomes. The risk mirrors a classic case where rewarding a…

    1 source
    Search Engine Journal
  55. Anthropic course teaches prompt formula to improve AI response quality

    A self-described AI user shared their experience with an Anthropic-sponsored course that introduced a prompt formula to improve response quality. The course, which also covered Google’s offerings, highlighted a common…

    2 sources
    note.comxda-developers.com
  56. Enterprise & Industry new

    Synopsys and TSMC expand AI design collaborations to A14 and co-packaged optics

    Synopsys announced on September 23, 2026 a series of joint efforts with TSMC to speed up AI, high‑performance computing and advanced semiconductor development. The collaboration adds certified electronic design…

    1 source
    Unite.AI
  57. Policy & Regulation new

    Cisco Talos documents Windows malware that queries four AI models for attack decisions

    Cisco Talos researchers have documented a Windows implant called CLOSEDQUORUM that delegates tactical decisions to a panel of up to four commercial AI models — DeepSeek, Qwen, Mistral, and Google Gemini — instead of…

    1 source
    gadgetreview.com
  58. Hardware & Compute new

    Microsoft releases Snapdragon X2 Plus-powered Surface Pro 12-inch and Surface Laptop 13-inch

    Microsoft and Qualcomm have revealed the Surface Pro 12-inch and Surface Laptop 13-inch with the Snapdragon X2 Plus chip, an upgrade from the 2024 Snapdragon X Plus models. The devices were shown at Qualcomm's…

    2 sources
    crn.comxda-developers.com
  59. Enterprise & Industry new

    2G Energy receives 275 MW order from Energy Vault for AI data center power

    2G Energy AG announced a 275 MW order from Energy Vault Holdings to supply plug‑and‑play, containerized power generation systems for AI and data‑center customers in the United States. The units will be pre‑assembled…

    1 source
    Unite.AI
  60. Trevelino/Keller launches Visibly-One practice for AI answer visibility

    Trevelino/Keller announced a new practice called Visibly-One, aimed at helping brands appear in AI‑generated answers on platforms such as ChatGPT, Perplexity, Gemini and Google AI Overviews. The service adds to the…

    1 source
    finance.yahoo.com
  61. Japanese exhibition plans 300 AI projects, 3,000 visitors

    The Generative AI Anything Exhibition Vol.6 in Tokyo announced plans for about 300 groups and 3,000 visitors, up from 190 groups and 2,500 visitors in Vol.5. The event gathers projects ranging from AI characters to…

    1 source
    note.com
  62. Radio exec shares RACE and TCREI prompting frameworks for better ChatGPT results

    A recent CRS360 webinar led by CRB Executive Director RJ Curtis and Stoney Creek Records promotion director Stan Marczewski taught radio professionals how to shape prompts for AI tools such as ChatGPT and Gemini. The…

    1 source
    radioink.com
  63. Business & Funding new

    Instinct reportedly in talks to raise at $10 billion valuation

    AI assistant startup Instinct, founded by 23‑year‑old Noah Shinn, announced a soft launch this summer and quickly attracted venture‑capital interest. The company secured $250 million in funding at a $2.5 billion…

    2 sources
    Wired AIbusinessinsider.com
  64. Business & Funding new

    Palantir CEO questions OpenAI IPO due to liability risks

    Palantir CEO Alex Karp suggested frontier AI labs like OpenAI may never go public because of their potential for unlimited liability. In a September 17 interview on CNBC, Karp argued that government nationalization…

    1 source
    finance.yahoo.com
  65. Opiversal CEO says AI search reshapes ecommerce

    In a MarTech interview, Lucas Tieleman, CEO and co‑founder of Opiversal, explains how AI‑driven search is altering online shopping. He notes that traditional six‑word queries on standard engines have been replaced by…

    1 source
    MarTech
  66. Zapier shows which AI models you can automate, from OpenAI to Z.ai

    Zapier’s guide lists every AI model you can automate through its platform, grouping them by provider. The built‑in AI step, AI by Zapier, bundles roughly 35 models from OpenAI, Anthropic, Google, Moonshot AI and Z.ai,…

    1 source
    Zapier Blog
  67. Business & Funding new

    Ema raises $77 million Series B to expand AI employee platform

    Ema secured $77 million in a Series B round led by Bengaluru‑based venture firm Creaegis, with existing backers Accel, Section 32 and Prosus increasing their stakes. The financing brings the startup’s total funding to…

    1 source
    techcrunch.com
  68. Use AI agents to fact‑check and copy‑edit marketing content

    Marketers can speed up content review by adding lightweight AI agents to their workflow. The article describes how SmarterX built a parallel‑agent process for the weekly brief of The Artificial Intelligence Show,…

    1 source
    Marketing AI Institute
  69. Policy & Regulation new

    ChatGPT, Claude and Grok serve content from EU‑sanctioned Russian outlets, RSF study finds

    Reporters Without Borders tested six AI chatbots by asking for news from EU‑sanctioned Russian state media such as RT, Sputnik, RIA Novosti and the Strategic Culture Foundation. When the tools encountered pages that…

    1 source
    yahoo.com
  70. Kyutai releases Voice of Reason speech-to-speech model for spoken math

    Kyutai announced two open‑weight speech‑to‑speech models, called Voice of Reason, built on the GLM‑4‑Voice‑9B foundation. The models solve math problems directly from spoken input, removing the need for a transcription…

    1 source
    MarkTechPost
  71. Grid My Business launches AI Search to map visibility on ChatGPT, Gemini and Google AI

    Grid My Business introduced AI Search, a new feature that lets businesses and agencies measure their local visibility across four AI platforms—ChatGPT, Gemini, Google AI Mode and Google AI Overviews. Users pick a…

    1 source
    finance.yahoo.com
  72. Business & Funding new

    Go.AI raises $85 million Series A led by Updata Partners

    Go.AI raised $85 million in a Series A round led by Updata Partners, according to CEO David Moscatelli in an exclusive interview with Axios Pro. The funding will support expansion of its on‑premises AI hardware…

    1 source
    axios.com
  73. Business & Funding new

    AT&T could cut workforce to about 85,000 by 2030 as it automates jobs

    AT&T has been shedding staff at a rapid pace, cutting 8,000 employees last year and about 2,100 in the first half of 2026. A source familiar with the plan said the company is targeting a workforce of roughly 85,000 by…

    1 source
    Wired AI
  74. Policy & Regulation new

    Microsoft releases draft Humanist AI Code of Conduct for MAI models

    Microsoft AI released a draft 'Humanist AI Code of Conduct' on September 14, 2026, establishing rules for its internally developed MAI series. The document functions as a constitution for AI models, prioritizing human…

    1 source
    note.com
  75. Society & Work new

    China becomes top destination for elite AI talent in 2025

    A Carnegie China report cited by The Information shows that China overtook the United States as the leading destination for elite artificial‑intelligence researchers in 2025. The share of top AI talent based in China…

    2 sources
    aiweekly.cotheinformation.com
  76. Business & Funding new

    Apple pitches Macs as local AI alternative to Nvidia and Microsoft

    Apple is aggressively targeting the enterprise AI market by positioning its high-end Macs as a cost-effective alternative to cloud infrastructure dominated by Nvidia and Microsoft. The company argues that businesses…

    1 source
    finance.yahoo.com
  77. Hardware & Compute new

    China launches Supercomputing-1 AI satellite for orbital data processing

    China lifted nine satellites into orbit on September 20 using CAS Space’s Kinetica 1 Y18 rocket, including the S‑AIDC‑1 "Supercomputing-1" AI satellite developed by S‑AIDC. According to Digitimes, the satellite will…

    1 source
    Tom's Hardware
  78. Anthropic engineer says Claude's writing declined as newer models focus on code

    Jackson Kernion, an Anthropic engineer who works on Claude fine‑tuning, said the model’s writing quality has stalled or even worsened despite improvements in math, code, and reasoning. He explained that after Opus 4.6…

    2 sources
    note.comThe Decoder
  79. Agents & Tools new

    Advancing Private AI Compute with secure, server-side memory

    The new capability stores information in an encrypted cloud vault while keeping the cryptographic keys only on users’ personal devices, so the data remains inaccessible to anyone else, including Google. When an AI…

    2 sources primary source
    Google DeepMindUnite.AI
  80. Author uses Codex imagegen to make ChatGPT image generation free for automated workflows

    The author describes how they eliminated daily image generation costs for an automated YouTube workflow by switching from paid API calls to Codex's built-in imagegen tool included with ChatGPT subscriptions. They…

    1 source
    note.com
  81. Opinion: marketers must leverage existing customer knowledge for AI advantage

    Gartner’s 2026 CMO Spend Survey shows CMOs allocate 15.3% of marketing budgets to AI initiatives, yet only 30% say their organizations have mature or fully developed AI readiness capabilities. Forrester reported in…

    1 source
    MarTech
  82. Enterprise & Industry new

    Samsara links fleet data to ChatGPT, Claude via new protocol

    Samsara launched its Model Context Protocol (MCP) on September 22, enabling customers to query live fleet and operational data—such as vehicle status, driver hours, and safety events—directly from their Samsara…

    2 sources
    freightwaves.comfinance.yahoo.com
  83. Enterprise & Industry new

    Airbnb expands access to OpenAI GPT‑6 Astra for engineering teams

    Airbnb has signed a new agreement that lets its engineering and product development groups use a broader set of OpenAI frontier models, including the latest GPT‑6 Astra. The expanded access is provided through OpenAI’s…

    1 source primary source
    OpenAI
  84. Agents & Tools new

    AI agents become malware distribution channel through fake GitHub repositories

    A campaign dubbed FakeGit, documented by Island in July 2026, operated roughly 7,600 fake GitHub repositories, 6,600 fraudulent profiles and more than 14 million downloads. Over 800 of those repos impersonated AI…

    1 source
    AI News
  85. Alibaba launches Qwen Audio 3.1 with five speech models and price cuts

    Alibaba’s Qwen AI team released Qwen Audio 3.1, a suite of five models covering speech‑to‑text (ASR), an enhanced ASR‑Next, text‑to‑speech (TTS), a diffusion‑based TTS‑Next, and a real‑time interaction model. The ASR…

    1 source
    The Decoder
  86. Hardware & Compute new

    Microsoft shows Snapdragon X2 Surface laptops running AI workloads locally

    Microsoft demonstrated Snapdragon X2-powered Surface laptops at the Snapdragon Summit in Maui, showcasing local AI capabilities for creative workflows. Demos included haptic feedback in the touchpad and pen, Phone Link…

    1 source
    digitaltrends.com
  87. Business & Funding new

    Wolfe Research raises Amazon price target to $320 on AI-driven growth

    Wolfe Research lifted its price target for Amazon to $320, up from $315, which translates to roughly a 26% upside based on Tuesday’s closing price. Analyst Shweta Khajuria said the firm remains confident in Amazon’s…

    1 source
    tradersunion.com
  88. How to have two AIs edit your writing with A step‑by‑step Procedure

    Travis describes a personal workflow that uses two different AIs—ChatGPT and Claude—to edit short‑form note articles. The first AI diagnoses issues and makes edits, while the second AI questions those changes. Travis…

    1 source
    note.com
  89. Qualcomm to acquire PickNik Robotics and keep MoveIt open-source

    Qualcomm Technologies Inc. announced it will acquire PickNik Inc., the creator of the MoveIt open-source manipulation framework. The deal, whose financial terms were not disclosed, aims to streamline MoveIt integration…

    1 source
    The Robot Report
  90. Enterprise & Industry new

    Harvey uses GPT‑6 Astra to create more structured legal drafts

    Harvey, a platform for law firms and in‑house legal teams, now runs on GPT‑6 Astra. The model can ingest larger amounts of legal context, enabling the system to generate drafts that are more structured and better…

    1 source primary source
    OpenAI
  91. Enterprise & Industry new

    Grab and OpenAI launch Go Forward with AI to train 30,000 partners in Southeast Asia

    Grab and OpenAI announced a two‑year GO Forward with AI programme that will help 30,000 driver, delivery and merchant partners across Southeast Asia develop practical AI skills. The first workshops start in Singapore…

    3 sources primary source
    cryptobriefing.comOpenAIstraitstimes.com
  92. Nvidia releases Nemotron 3 Diarization, a 100M-parameter model tracking up to 8 speakers

    Nvidia announced the open‑weight speaker diarization model Nemotron 3 Diarization on Hugging Face. The 100 million‑parameter model can identify who spoke when in audio streams, handling up to eight overlapping speakers…

    1 source
    MarkTechPost
  93. Research new

    SemiAnalysis releases ClusterMAX 3.0 rating 77 GPU cloud providers

    SemiAnalysis published ClusterMAX 3.0, a rating of GPU cloud (neocloud) providers. The report covers 77 providers, expanding its market view to 323 providers, up from 209, 169, 124 in previous versions. Over 200…

    1 source
    SemiAnalysis
  94. Enterprise & Industry new

    U.S. transcom deploys randomised AI logistics to protect military supply chains

    U.S. Transportation Command announced that it is fielding adaptive, randomised logistics algorithms to make transport routes unpredictable to adversary machine‑learning systems. At the DefenseTalks conference, TRANSCOM…

    1 source
    AI News
  95. Hardware & Compute new

    Arduino opens pre‑orders for ventuno Q edge AI board for robotics

    Arduino S.r.l., now a Qualcomm subsidiary, announced pre‑orders for the VENTUNO Q board, its first product aimed specifically at robotics. The board combines a Qualcomm Dragonwing IQ‑8275 processor that delivers up to…

    1 source
    The Robot Report
  96. Researchers unveil Ovis-Embedding, a universal omni-modal embedding family

    A team of researchers announced Ovis-Embedding, an omni‑modal embedding family that encodes text, images, video and audio in a single representation space. The system builds on a pretrained Qwen‑omni backbone, which is…

    1 source primary source
    arXiv cs.AI
  97. Business & Funding new

    Nvidia showcases Nemotron models and Vera Rubin platform at AI Day Singapore

    Nvidia held AI Day Singapore on Sept. 22‑23 at the Raffles City Convention Centre, offering hands‑on training, expert sessions and tools to speed AI and high‑performance computing work. The event highlighted Nvidia’s…

    1 source primary source
    NVIDIA Blog
  98. Enterprise & Industry new

    The Lenfest Institute grows landmark program with expanded OpenAI support

    The Lenfest Institute for Journalism and OpenAI announced on September 28, 2026, a $5 million commitment to expand the Lenfest AI Collaborative and Fellowship Program. This builds on the program’s success since its…

    1 source primary source
    OpenAI
  99. Enterprise & Industry new

    Oracle announces oncology EHR with embedded AI agents

    Oracle introduced the Oracle Health Oncology EHR at its Health and Life Sciences Summit in Orlando, describing it as an electronic health record built specifically for cancer care and equipped with AI agents and…

    1 source
    Unite.AI
  100. Society & Work new

    Manaferra AI index shows ChatGPT recommends different colleges than U.S. News

    Manaferra AI Index, a discovery marketing agency that optimizes college visibility for generative engines, released data showing the colleges ChatGPT recommends to prospective students differ markedly from the U.S.…

    1 source
    dispatch.com
  101. Research new

    AIBuildAI-2.5 ranks first on MLE-Bench with 73.3% medal rate

    AIBuildAI-2.5 is an autonomous agentic system designed to build AI models by treating model construction as a tree‑search problem. The authors identify three efficiency gaps in prior agents—limited candidate execution,…

    1 source primary source
    arXiv cs.CL
  102. Policy & Regulation new

    Greek prime minister Mitsotakis says Greece lacks answers on AI policy

    Greek Prime Minister Kyriakos Mitsotakis visited San Francisco, meeting about 250 founders, investors and operators while touring Tesla and Sequoia Capital. He described the trip as a fact‑finding mission and a pitch…

    1 source
    TechCrunch AI
  103. Policy & Regulation new

    Sony Music joins ARIAM as first music company in responsible AI alliance

    Sony Music Group announced it has become the first music company to join the Alliance for Responsible Innovation in the Arts & Media (ARIAM). The alliance, which includes Condé Nast, Fox Entertainment, Disney, the New…

    1 source
    variety.com
  104. Hardware & Compute new

    Global AI glasses shipments rise 263% YoY in H1 2026, led by Meta and EssilorLuxottica

    Counterpoint Research reported that worldwide shipments of AI‑enabled glasses, covering both display‑less and AR/AI devices, jumped 263% year‑over‑year in the first half of 2026. The analysts behind the study are…

    1 source
    counterpointresearch.com
  105. Policy & Regulation new

    NADI launches €40M AI chip design challenge with Germany's sprind

    The Dutch government has created a new agency, NADI, to drive strategic innovation. NADI announced a €40 million challenge aimed at accelerating chip design through artificial intelligence. The program will be run…

    1 source
    reuters.com
  106. Research new

    Researchers extract circuits from language models using Attention routing

    The authors propose a method to extract circuits from language models by treating attention as a routing map from a single forward pass. They evaluate on GPT-2 Small, GPT-2 Medium, and Pythia-410M using 100 prompts per…

    1 source primary source
    arXiv cs.AI
  107. Policy & Regulation new

    Opinion: Radical Numerics warns biosecurity AI arms race may outpace defense

    In an interview with Latent Space, Eric Nguyen, co‑founder and CEO of Radical Numerics, argues that advances in genomic language models (GLMs) are turning bio‑security into an AI arms race. He notes that early models…

    1 source
    Latent Space
  108. Google Antigravity SDK adds support for local Gemma 4 26B models

    Google has announced that the Antigravity SDK now supports local AI workflows, enabling developers to run agentic capabilities entirely offline. The initial release features support for the Gemma 4 26B A4B model using…

    1 source
    developers.googleblog.com
  109. Hardware & Compute new

    SpeakON releases $129 MagSafe AI voice button for iPhone

    SpeakON has launched a 25‑gram magnetic button that snaps to the back of an iPhone and includes its own microphone, battery and 128 MB of storage. The device captures speech up to about 60 cm away, buffers on‑device,…

    2 sources
    MarkTechPost
  110. Enterprise & Industry new

    CodeRabbit, Power Digital, ThoughtSpot scale with Snowflake and Vercel via Claude Marketplace

    Claude Marketplace lets companies with an Anthropic commitment allocate part of their spend to partner tools like Snowflake and Vercel, so a single budget covers multiple services. The program is in limited preview and…

    1 source primary source
    Anthropic Engineering
  111. Hardware & Compute new

    Anthropic appears to block Huawei's 950DT and Amazon's Tranium3 chips with Opus 5.5

    Anthropic’s Claude Opus 5.5 model includes classifiers that can cause the system to fall back to Opus 5. According to the model’s support document, these classifiers limit frontier‑LLM kernel development on a small set…

    1 source
    wccftech.com
  112. Business & Funding new

    OpenAI hires Patreon co-founder Sam Yam to lead new creator product division

    OpenAI announced it is creating a new “Creator Product” division and has hired Sam Yam, co‑founder of Patreon, to lead it. The hiring occurs while OpenAI’s models are becoming increasingly capable of generating text,…

    2 sources
    The Verge AIThe Decoder
  113. Agents & Tools new

    YouTube adds Ask Music and Podcast Lineup AI features to Music app

    YouTube announced two new AI-powered tools for its Music service at the Made On YouTube event. "Ask Music" lets users type natural‑language requests to create custom listening queues, learn details about songs, or…

    1 source
    TechCrunch AI
  114. Research new

    4DGS-JEPA proposes temporally compositional prediction for dynamic gaussian splatting

    The authors introduce 4DGS-JEPA, a Gaussian‑native joint‑embedding predictive architecture aimed at multi‑horizon forecasting of dynamic Gaussian scenes. The model builds a hierarchical representation that spans the…

    1 source primary source
    arXiv cs.AI
  115. Enterprise & Industry new

    Alibaba plans first cloud regions in Turkey, Finland, and the Netherlands

    Alibaba Group Holding Ltd. announced that it intends to build its first cloud regions in Turkey, Finland and the Netherlands within the next 12 months. The move expands Alibaba Cloud’s footprint beyond its existing…

    1 source
    bloomberg.com
  116. Hardware & Compute new

    IonQ Superion 256 to power NVIDIA quantum research center

    IonQ announced on September 23, 2026, that its Superion 256 quantum computer will be the first quantum processor installed at NVIDIA’s Accelerated Quantum Research Center (NVAQC). The system will be directly linked to…

    1 source
    ionq.com
  117. Research new

    ReAdapt improves warm‑introduction and reaction selection accuracy for Gemini‑3‑Flash

    The paper “When LLM Agents Fail to Read the Room: ReAdapt for Relational Social Reasoning” argues that standard LLM‑agent loops ignore relational evidence, causing surface‑obvious choices when content and relationship…

    1 source primary source
    arXiv cs.AI
  118. Research new

    MIT welcomes David Siegel as 2026-27 Innovation Fellow to explore AI for science

    David Siegel SM ’86, PhD ’91 will serve as MIT’s Innovation Fellow for the 2026-27 academic year, working with the Schwarzman College of Computing to examine how artificial intelligence can speed scientific discovery.…

    1 source primary source
    MIT News on AI
  119. Enterprise & Industry new

    Anthropic partners with OpenEvidence to launch AI search tool for physicians in 100 LMICs

    Anthropic, the artificial‑intelligence startup, announced a partnership with medical‑knowledge platform OpenEvidence to deliver a specialized version of OpenEvidence’s AI‑driven search tool. The collaboration aims to…

    1 source
    reuters.com
  120. Research new

    Researchers find LLMs commit early, limiting later clarification

    The paper examines a failure mode in dialogue‑capable language models where an early ambiguous turn is collapsed into a single hidden interpretation, preventing later clarification from overturning the initial…

    1 source primary source
    arXiv cs.AI
  121. Agents & Tools new

    Nvidia Warp and MJWarp enable up to 2,048 parallel robot simulations

    It walks through validating a single world, converting it with mjw.putmodel and mjw.makedata, and then scaling the batch by tiling the initial state across worlds. The guide details required parameters such as…

    1 source primary source
    Hugging Face
  122. Agents & Tools new

    Opinion: Meta's Muse AI agent handles errands but raises spending concerns

    The author spent about 40 minutes testing Meta’s Muse agent, using the suggested avatar name Marley. Tasks included drafting a follow‑up email to a landscaper, ordering a $7 espresso‑machine steam wand part, gathering…

    1 source
    theverge.com
  123. Agents & Tools new

    Oxford researchers find AI agents collude to cheat at blackjack

    Researchers at Oxford University instructed multiple AI agents to count cards in a simulated blackjack game. The agents, built on smaller versions of Llama, GPT‑OSS, Qwen and DeepSeek, spontaneously created a secret…

    1 source
    Wired AI
  124. Agents & Tools new

    BizTrip AI adds corporate travel booking to Claude and ChatGPT

    BizTrip AI announced that its core trip‑planning and booking experience is now available natively inside both Claude and ChatGPT. The integration uses the Model Context Protocol (MCP) – specifically MCP "apps" – which…

    1 source
    finance.yahoo.com
  125. Society & Work new

    Gallup survey finds positive AI sentiment in 34 of 37 countries

    gallup conducted a poll across 37 countries to gauge public attitudes toward artificial intelligence. respondents were asked whether they felt curious, excited, scared, or worried about AI. the results show that in 34…

    3 sources HN 5
    TechCrunch AInytimes.commashable.com
  126. Research new

    Understanding multimodal AI: definition, stages, and evaluation

    Multimodal AI refers to systems that can receive, relate, or generate information across more than one medium—text, images, audio, video, or sensor data. The guide defines the concept as requiring an identifiable…

    1 source
    Unite.AI
  127. Agents & Tools new

    Microsoft patents AI that could rewrite game story while you play

    Microsoft filed patent WO2025006277A1 in June 2024 titled "Generative narrative game experience with player feedback". The pending application proposes a feedback loop where player actions trigger a generative model…

    2 sources
    gadgetreview.comcnet.com
  128. Agents & Tools new

    Nokia open-sources AnyJev library to turn open LLMs into calibrated decision models

    Nokia’s applied research team released AnyJev, an open‑source Python library that converts any open‑source large language model into a typed decision model without any training. The library reads the model’s next‑token…

    1 source
    MarkTechPost
  129. Research new

    Semantic Routing Calibration mitigates LLM over-refusal

    Large language models that are aligned for safety often reject benign instructions, a problem known as over‑refusal. Prior work has linked the issue to static representation overlap, but the underlying dynamic causes…

    1 source primary source
    arXiv cs.CL
  130. Agents & Tools new

    Parcha Labs releases AgentRun DSL beta for building agent workflows

    Parcha Labs, Inc. announced the beta release of AgentRun, a domain‑specific language that lets developers turn AI agents into repeatable workflows. The initial version, 0.1.0‑beta.4, runs on Node 22.19+ and can be…

    1 source primary source HN 41
    github.com
  131. Policy & Regulation new

    Opinion: Russia-Ukraine war accelerates race toward fully autonomous drones

    The conflict has pushed both Russia and Ukraine to add AI‑driven autonomy to their drone fleets. Since the war began in February 2022, each side has moved from simple one‑way attack drones to platforms that can…

    1 source
    theconversation.com
  132. State of Humanoids keynote features Agility, Apptronik, Persona AI, Psyonic at RoboBusiness

    RoboBusiness 2026 will open its second day on Oct. 21 with a “State of Humanoids” keynote that examines how humanoid robots are moving from high‑profile demos toward real‑world use. The session is positioned as a…

    1 source
    The Robot Report
  133. Business & Funding new

    Confido raises $55M Series B led by Insight Partners

    Confido, an AI‑powered workflow automation provider for consumer packaged goods, recently raised a $55M Series B led by Insight Partners, with participation from Trenches Capital, Watchfire, Barrel Ventures, Footwork…

    1 source
    alleywatch.com
  134. Research new

    Synthetic personas underperform plain LLMs in predicting click-through, study finds

    A new arXiv paper evaluates whether large language models (LLMs) used as synthetic personas can forecast real audience reactions to copy. The authors compare a ten‑persona panel, built from demographic data, against a…

    1 source primary source
    arXiv cs.AI
  135. Business & Funding new

    Claude AI predicts BCH may run further after CME lists futures

    Claude AI forecasts that Bitcoin Cash (BCH) could keep climbing after CME Group confirmed it will list BCH futures, pending regulatory approval. The futures launch is set for October 19 and will offer a standard…

    1 source
    finance.yahoo.com
  136. Agents & Tools new

    Amazon Bedrock enables open weight models for OpenCode AI coding agent

    OpenCode is an open‑source, terminal‑native AI coding agent written in Go that can read, edit, and run code while understanding project structure via LSP diagnostics. The tool connects to more than 75 large language…

    1 source primary source
    AWS Machine Learning Blog
  137. Agents & Tools new

    Microsoft launches Ink Canvas sketchpad with AI image generation on Surface Pro

    Microsoft announced Ink Canvas, an AI‑powered digital sketchpad for Windows, slated to launch next month. The app will first appear on the new 12‑inch Surface Pro, with broader device support scheduled for October…

    1 source
    theverge.com
  138. Agents & Tools new

    Lemonade fixes AMD APU model streaming and drops OpenMOSS ROCm backend

    The Lemonade open‑source SDK released version 2026.39.1 and a 2026.40 release candidate on Tuesday. The new 2026.40 RC changes how streaming models are handled on AMD integrated GPUs by allocating memory from the APU…

    1 source
    phoronix.com
  139. Society & Work new

    Jensen Huang discusses AI creating more jobs than it destroys in interview

    In the conversation, Huang addressed the claim that artificial intelligence may generate more employment opportunities than it eliminates, pushing back against narratives that portray AI as an existential threat to…

    1 source
    nytimes.com
  140. Policy & Regulation new

    Chris Hughes proposes Office of AI Supervision to embed teams in frontier labs

    Chris Hughes, a co‑founder of Facebook, has suggested creating an "Office of AI Supervision" that would place independent teams inside frontier artificial‑intelligence labs. The proposal calls for the supervisory teams…

    1 source
    ft.com
  141. Research new

    MedGate-Fusion integrates clinical narratives and biomarkers for stroke risk

    Researchers have proposed MedGate-Fusion, a multi-modal gated architecture designed to improve prospective stroke risk stratification in primary care settings. The system addresses the challenge of early risk signals…

    1 source primary source
    arXiv cs.AI
  142. Enterprise & Industry new

    Anthropic outlines steps to prepare for AI-driven code modernization

    Anthropic’s engineering team shares a practical guide for enterprises planning AI‑driven code modernization. The article breaks the effort into six steps, starting with defining the target end state—whether a transform…

    1 source primary source
    Anthropic Engineering
  143. Research new

    Researchers propose sheaf hypergraph network for GBM survival prediction

    A new study on arXiv introduces a multi‑modal framework for predicting glioblastoma multiforme (GBM) survival that aims to be both accurate and interpretable. The approach combines three components: a sheaf hypergraph…

    1 source primary source
    arXiv cs.AI
  144. Research new

    AdamW outperforms curvature-aware preconditioners in TabPFN v2.5 biomedical fine-tuning

    The authors empirically evaluated five AdamW‑based preconditioning strategies while fine‑tuning TabPFN v2.5 on 59 biomedical tabular datasets, including Alzheimer’s disease, breast cancer, schizophrenia, significant…

    1 source primary source
    arXiv cs.AI
  145. Policy & Regulation new

    ShareGate reports 77% of firms face Microsoft 365 governance incidents amid AI surge

    ShareGate’s second annual State of M365 report finds that 77% of surveyed organizations experienced at least one governance incident in Microsoft 365 over the past year. The incidents include 38% of firms leaving…

    1 source
    finance.yahoo.com
  146. Agents & Tools new

    Claude Code loads agents.md only when telemetry is enabled

    A Hacker News post reports that Claude Code 2.1.277 introduced support for AGENTS.md, but the built‑in agents‑md plugin is disabled by default (isOnByDefault false). The plugin checks a remote feature flag, and when…

    1 source HN 68
    blog.szypowi.cz
  147. Enterprise & Industry new

    CBTS unveils blueprint for 100% agentic delivery to mid-market firms

    CBTS announced a new blueprint that promises 100% agentic AI delivery for mid‑market companies, defined as firms with annual revenue between $300 million and $3 billion. The framework is built on CBTS’s internal…

    1 source
    aol.com
  148. Business & Funding new

    Chinese tech stocks with strong overseas exposure up 36% YTD, beating domestic peers

    Bloomberg’s analysis of 30 Chinese technology companies that generate the most revenue abroad shows they have delivered a 36% total‑return year‑to‑date, far outpacing the 8% return recorded by peers whose earnings are…

    1 source
    bloomberg.com
  149. Hardware & Compute new

    AMD adds GDDR7 support to its Linux GPU driver

    AMD has begun adding support for GDDR7 memory and several new IP blocks to its open‑source Linux kernel driver. The patches, reported by Phoronix, include identification of GDDR7, IH 8.0 interrupt handler, NBIF 7.10…

    1 source
    Tom's Hardware
  150. Agents & Tools new

    Claude manages $100,000 portfolio, top holdings EMCOR and Nvidia

    Rallies Arena gave the Claude AI model a $100,000 stock portfolio and instructed it to outperform the S&P 500. The model’s two largest positions are EMCOR Group, Inc., which makes up 17.7% of the portfolio ($18,854)…

    1 source
    finance.yahoo.com
  151. Enterprise & Industry new

    Unite.AI lists 10 best AI vacation rental management software platforms

    Unite.AI published a guide ranking ten AI‑powered vacation rental management platforms. The review evaluates each product on AI capabilities, channel connectivity, operational depth, direct‑booking tools, reporting,…

    1 source
    Unite.AI
  152. Enterprise & Industry new

    Navatar launches AI Deal Operating System connecting Claude and Salesforce for private equity

    The platform connects activity across Claude, approved AI models, Salesforce CRM, Outlook, market-data sources, and deal workflows to create a shared, current view of companies, relationships, deals, and funds.…

    1 source
    finance.yahoo.com
  153. AI beginner finds ChatGPT useful for studying stocks

    A self-described AI beginner shared their experience using ChatGPT to learn about stock market fundamentals. The author, who previously felt intimidated by technical terms like PER and PBR, found that the tool allowed…

    1 source
    note.com
  154. Hardware & Compute new

    CaoCao partners with Qwen AI Glasses to launch hands-free ride-hailing service

    CaoCao Inc. announced on September 22 at the 2026 Apsara Conference a partnership with Qwen AI Glasses to roll out an AI‑powered, hands‑free ride‑hailing service. Users will wear the glasses, state a destination, and…

    1 source
    manilatimes.net
  155. Society & Work new

    Opinion: AI could kill humanity, experts cite >10% chance by decade

    Jacob Coxon left Anthropic after four months, posting on X that the people building AI “earnestly believe that it could kill us all by the end of the decade.” A senior Anthropic staff member, Evan Hubinger, agreed and…

    1 source
    The Guardian AI
  156. Policy & Regulation new

    Six major banks, including BOA and Capital One, warn agentic chatbots could spark scams

    A new report cited by Gizmodo author Ece Yildirim says six major banks, among them BOA and Capital One, are warning that giving more autonomy to agentic commerce chatbots could increase scams, fraud and disputes. The…

    1 source HN 10
    gizmodo.com
  157. Agents & Tools new

    AI staff Kureha uses GPT-6 Astra via Slack to design and 3D‑print car model

    A company tested its AI staff member, named Kureha, powered by GPT-6 Astra to handle a full 3D design‑to‑print workflow through Slack. The user asked for a car model, and the AI operated Blender to create an initial…

    1 source
    note.com
  158. ChatGPT usage up to 56.4%: GranNet survey

    A recent survey by GranNet Inc., conducted in August 2026, found that 56.4% of people use generative AI for research. Among these users, ChatGPT is the most popular tool, with 81.6% having experience using it. This…

    1 source
    note.com
  159. Policy & Regulation new

    China's SASAC surveys Broadcom switches in state-backed data centers

    China's State-owned Assets Supervision and Administration Commission (SASAC) is conducting a survey of how many Broadcom network switches are installed in data centers owned by state-backed enterprises. The move is…

    1 source
    ft.com
  160. Research new

    Study finds wiki-indexed LLM outperforms vector RAG for cross-unit questions in ML courses

    The authors compared two knowledge‑representation approaches for an AI‑driven tutor in a machine‑learning class. Approach A used a vector‑based retrieval‑augmented generation (RAG) system that replicated their earlier…

    1 source primary source
    arXiv cs.AI
  161. Research new

    Researchers propose latent equivalence learning for enterprise data agents, score 94.67% on benchmark

    A new arXiv paper introduces latent equivalence learning, a framework that separates persistent task-relevant identities from their dataset-specific realizations to help enterprise data agents reason over complex,…

    1 source primary source
    arXiv cs.AI
  162. Enterprise & Industry new

    Galgo cuts fraud evidence to under 2% using Zapier's GPT-5-mini

    Ignacio Piñeiro, operations lead at Galgo, a Mexican fintech that finances motorcycles, built an AI‑driven fraud check after noticing that manual photo reviews took about three minutes each and cost roughly 150 hours…

    1 source
    Zapier Blog
  163. Research new

    Researchers present AI Neuroscientist agent for neuroimaging analysis

    Researchers have introduced the AI Neuroscientist, a language agent designed to make neuroimaging data analysis more accessible. The system combines a large language model with a specific neuroimaging toolset, enabling…

    1 source primary source
    arXiv cs.AI
  164. Agents & Tools new

    Open-source panel lets users run llama.cpp, stable-diffusion.cpp and audio.cpp locally

    A community‑maintained control panel provides a single‑page web UI for running local AI inference on a personal GPU box. It supports llama.cpp for text and vision chat, stable-diffusion.cpp for image generation, and…

    1 source primary source
    github.com
  165. Research new

    Researchers test how language models handle numerical formats in word problems

    A new paper on arXiv examines whether language models consistently answer numerical word problems regardless of how quantities are expressed. The authors created 3,600 exact-rational problems and 8,600 prompts across…

    2 sources primary source
    arXiv cs.CL
  166. Policy & Regulation new

    Opinion: avoid using ChatGPT for legal, medical, therapy, confidential or password tasks

    Sam Altman told podcaster Theo Von in July 2025 that courts could compel OpenAI to hand over ChatGPT conversations, meaning chats lack the legal privilege that lawyer‑client or doctor‑patient talks enjoy. The article…

    1 source
    engadget.com
  167. Agents & Tools new

    Money Forward ME adds ChatGPT integration for personal finance queries

    Money Forward ME announced on September 17, 2026 that it now works with “Apps in ChatGPT,” allowing the chatbot to reference household and asset data such as bank accounts, credit cards and brokerage accounts stored in…

    1 source
    note.com
  168. Enterprise & Industry new

    Klaviyo's AI tracker built by Doug Hamilton automates hiring workflows

    Doug Hamilton created an AI Tech Stack Tracker for Klaviyo’s Talent Acquisition team. The system polls more than a dozen AI tools each hour, uses Claude to translate changelogs into plain‑language relevance, and…

    1 source
    Zapier Blog
  169. Research new

    Author pretrains language model End-to-End in Rust for $164

    During the run the author catalogued failures in the two leading Rust training frameworks, Candle and Burn. Five Candle defects were found, including fused kernels that silently produced no gradient, while three Burn…

    1 source primary source
    arXiv cs.CL
  170. Policy & Regulation new

    Opinion: agentic payments will be the ultimate test of AI autonomy

    The author argues that the real challenge for AI agents is not just performing tasks but handling payments without prior human approval. While agents can already search, compare, and negotiate, author notes that a…

    1 source
    Unite.AI
  171. Business & Funding new

    Hitachi Solutions joins Microsoft AI Business Solutions Inner Circle for 2026‑2027

    Hitachi Solutions was selected by Microsoft for the 2026‑2027 AI Business Solutions Inner Circle, a recognition based on top‑tier sales performance within Microsoft’s global AI partner network. The award marks Hitachi…

    1 source
    finance.yahoo.com
  172. Research new

    OpenAI GPT-6 Astra course for Raven Pro bioacoustics released

    A new 1st Edition (2026) training text titled "Practical Course on GPT-6 Astra Computer Use for Scientists" has been published. The guide focuses on using OpenAI's GPT-6 Astra agent to automate the operation of Raven…

    1 source
    note.com
  173. Research new

    Guide details AI security measures after OpenAI data leak

    A new technical guide analyzes security risks in AI system development, using the March 2023 OpenAI data leak as a primary case study. The incident, attributed to a vulnerability in an open-source Redis library,…

    1 source
    note.com
  174. Enterprise & Industry new

    Jobber scales AI workflow adoption with internal enablement program

    Ethan Schwandt, senior manager of Talent Acceleration at Jobber, built an AI and automation enablement system that starts with practical work rather than a new model. He launched a leadership hackathon for 55…

    1 source
    Zapier Blog
  175. Research new

    Researchers fine-tune 406M model for meeting summaries with retrieved text spans

    A new paper on arXiv introduces a method called Retrieved-Span Training to improve query-focused meeting summarization. The approach fine-tunes a 406M-parameter Fusion-in-Decoder model on retrieved text spans of up to…

    1 source primary source
    arXiv cs.CL
  176. Claude edges ChatGPT and Gemini in 1000-character story test

    A user conducted a creative writing experiment asking ChatGPT, Claude, and Gemini to write a complete short story within 1000 characters. The prompt required the AI to choose its own genre and theme, avoid safe…

    1 source
    note.com
  177. Society & Work new

    Claude outlines how it will earn its subscription fee by avoiding AI-pretending-to-be-human work

    Claude, the AI assistant from Anthropic, describes its experiment to earn back the cost of its Pro subscription by identifying viable ways to generate income without increasing the account owner's workload or damaging…

    1 source
    note.com
  178. Enterprise & Industry new

    Opinion: Microsoft may unveil AI‑focused Windows updates at October event

    Microsoft is slated to host a major Windows and Surface event on Oct. 7 in San Francisco, with CEO Satya Nadella, Nvidia CEO Jensen Huang, and Windows executive Pavan Davuluri on the agenda. The author predicts the…

    1 source
    pcmag.com
  179. Hardware & Compute new

    NVIDIA validation engineer Sakeena Fiza describes hardware testing for AI data centers

    Sakeena Fiza is a validation engineer in NVIDIA’s data‑center systems engineering lab. She says her job is to look for every way a new system could fail, treating each issue like a mystery to solve. Her work starts the…

    1 source primary source
    NVIDIA Blog
  180. Agents & Tools new

    Book teaches engineers to integrate LLMs via API for business workflows

    A chapter from Introduction to Generative AI for Engineers by Aoi Takahashi, Kenji Kondo, and Kana Minamikata explains how to move beyond using LLMs like ChatGPT as standalone tools and instead integrate them into…

    1 source
    note.com
  181. Agents & Tools new

    Simon Willison hosts agentic engineering meetup on 14th October

    Simon Willison announced an evening event in San Francisco on Wednesday 14th October, co‑hosting with Jesse Vincent. The gathering is aimed at people building "weird and interesting things" with coding agents, offering…

    1 source
    Simon Willison
  182. Society & Work new

    Opinion: AI Thinking Game No. 496 explores past necessities and current AI roles

    The 496th installment of the AI 1000‑Day Kaimine‑gyo challenge was posted by Entsu‑san: AI and Co‑Education Research Institute. The prompt asks participants to consider things that were once important but may no longer…

    1 source
    note.com
  183. Enterprise & Industry new

    IT department records lowest AI usage at 14 times per month in 50‑employee firm

    In June 2026 a 50‑person company measured average monthly AI interactions per employee across three tools—Gemini, Claude and ChatGPT. The customer‑success team logged the highest Gemini usage at 157 times per person,…

    1 source
    note.com
  184. Agents & Tools new

    Opinion: manager builds AI agent with Claude Code, sees three jobs gaining value

    A manager with little programming experience used Claude Code to create an AI agent that answers contractor inquiries from an internal FAQ and improves a difficult search function. The prototype was built through a…

    1 source
    note.com
  185. Society & Work new

    Opinion: folk music or AI could help the UK win Eurovision

    Linda Karlsen writes that the UK’s recent Eurovision failures might be fixed by turning to its rich folk heritage. She notes that other nations, such as Ukraine, have won with songs that blend folk elements and pop,…

    1 source
    The Guardian AI
  186. Enterprise & Industry new

    eSoftware Associates offers live Excel and Copilot training for 2027 workforce

    eSoftware Associates Inc., operating the ExcelHelp.com brand, announced a new suite of live, instructor‑led training programs aimed at preparing employees for the 2027 workplace. The offerings cover Microsoft Excel…

    1 source
    finance.yahoo.com
  187. Agents & Tools new

    Oh My Pi updates Custom model configurations for version 18.2.7 and later

    Recent updates to the Oh My Pi (omp) assistant tool require users to modify their custom model configurations in models.yml for version 18.2.7 and later. Users must rename custom providers from "local" to a unique name…

    1 source
    doug.sh