AI Inside Daily (for Patrons)

Folgen

• 
  AI INSIDE DAILY | Friday, July 31, 2026 31.07.2026 9Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:Anthropic disclosed yesterday that three of its Claude models gained unauthorized access to real production systems at three different organizations during cybersecurity evaluations that should have been sealed off from the internet. Claude Opus 4.7, Mythos 5, and an internal research model all breached real systems through a misconfiguration at Anthropic's third-party partner Irregular. Only the newest model stopped on its own.https://www.anthropic.com/news/investigating-incidents-cybersecurity-evalsAmazon reported after the bell yesterday. AWS grew 37 percent, its fastest quarter in more than four years, and Andy Jassy said AWS's AI and custom-chips businesses have each crossed 25 billion dollars in annual run rate.https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-aws-revenue-growth-q2-2026-earningsReuters reviewed more than 80 Chinese academic papers and patents and found military-linked researchers using OpenAI's GPT-3.5 and Anthropic's Claude 3 Haiku to train domestic defense AI. One PLA cyber-warfare unit distilled GPT-3.5 to summarize classified military code.https://www.defensenews.com/industry/techwatch/2026/07/31/chinese-military-researchers-tap-us-ai-models-to-train-defense-systems/Thinking Machines Lab released Inkling-Small, a 276-billion-parameter open-weights model with 12 billion active parameters, on Hugging Face. It beats its own bigger Inkling sibling on some agentic benchmarks at one-quarter the size. Same day the company lost a co-founder to OpenAI.https://thinkingmachines.ai/news/inkling-small/DeepSeek pushed the public beta of its V4-Flash API this morning as V4-Flash-0731. Same architecture as the preview, retrained, with Terminal Bench 2.1 jumping from 61 to 82. Now beats DeepSeek's own V4-Pro-Preview on several agent coding benchmarks, and anyone already calling V4-Flash got silently upgraded.https://api-docs.deepseek.com/updates/The Munich Regional Court ruled this morning that Suno violated copyright by memorizing and reproducing six GEMA-represented songs, including tracks by 80s band Alphaville. First major EU ruling saying generative AI music companies must license their training catalogs.https://variety.com/2026/digital/news/suno-loses-ai-lawsuit-gema-1236825010/Google rolled back Nano Banana image generation inside Google Earth less than 24 hours after launching it, after reporters showed the tool could fabricate satellite images of a Los Angeles blast crater and protesters outside Google's own Mountain View campus.https://www.engadget.com/2228142/google-rolls-back-the-needless-ai-generation-tools-it-added-to-google-earth/Patrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI INSIDE DAILY | Friday, July 24, 2026 24.07.2026 13Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:Anthropic launched Claude Opus 5. It approaches Fable 5 on coding and knowledge work at half the price, becomes the default on Claude Max, and adds a low/medium/high effort dial. Anthropic also says it's the most aligned model they've ever shipped on their internal behavioral audit.https://www.anthropic.com/news/claude-opus-5Twenty-five tech companies including Nvidia, Microsoft, Meta, IBM, Dell, Palantir, a16z, Mistral, Hugging Face, and Y Combinator signed a joint "Open Weights and American AI Leadership" letter. OpenAI, Anthropic, and Google notably did not sign. Jensen Huang used his first-ever X post to push it.https://www.benzinga.com/markets/tech/26/07/60673099/meta-microsoft-palantir-nvidia-and-21-others-sign-letter-backing-open-weight-ai-modelsThree more shoes dropped on the Meta glasses harassment story: Monopoly Events banned Meta glasses from its UK Comic-Con shows, Scottish ferry operator CalMac paused bridge visits after a passenger secretly filmed crew, and Engadget reported the glasses have picked up the public nickname "pervert glasses."https://www.engadget.com/2222008/instagram-is-now-banning-users-who-make-creepy-content-with-meta-glasses/Midjourney announced its first-ever acquisition and it is not what you would guess. It bought Co-Star, the astrology app. Founder Banu Guler joins as Design Director across all Midjourney projects. David Holz says half a dozen more ambitious projects are coming in the next six months.https://updates.midjourney.com/midjourneys-first-acquisition/Google expanded Gemini Spark, its agentic AI assistant, to $20/month AI Pro subscribers in the US. Previously it was locked behind Ultra at $100 to $200 a month. Spark runs on Gemini 3.5 and plugs into Gmail, Calendar, Docs, Sheets, and Slides.https://www.engadget.com/2222303/google-expanding-access-gemini-spark/Patrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI Inside Daily | Friday, July 17, 2026 17.07.2026 11Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:The 2026 World AI Conference opened in Shanghai Friday and Chinese President Xi Jinping delivered the keynote in person, his first WAIC appearance since the event started in 2018; Xi formally launched the World Artificial Intelligence Cooperation Organization (WAICO), a 29-nation intergovernmental AI alliance headquartered in Shanghai with founding members including Russia, Brazil, Indonesia, Malaysia, South Africa, Senegal, Pakistan, and Cuba; Xi said AI development "should not be a solo performance by a single country, but a symphony of international cooperation" and warned against "overstretching the national security concept" in an apparent jab at Washington; also pledged 5,000 AI training slots for developing countries over five years and UN Secretary-General Antonio Guterres attended in the room.https://www.aljazeera.com/news/2026/7/17/chinas-xi-jinping-launches-new-ai-alliance-what-is-itMoonshot AI officially released Kimi K3 Thursday, a 2.8-trillion-parameter multimodal mixture-of-experts model with a 1-million-token context window, priced at $3 per million input tokens and $15 per million output tokens (Claude Sonnet 5 tier); full open weights ship on Hugging Face on July 27th; on Artificial Analysis's Intelligence Index K3 scored a 57 and ranks fourth of 189 models tested; Moonshot's own benchmarks put K3 behind Anthropic Fable 5 and OpenAI GPT-5.6 Sol but ahead of Claude Opus 4.8, GPT-5.5, Sonnet 5, and GLM-5.2, positioning Chinese open weights as Opus-class performance at Sonnet-tier pricing.https://venturebeat.com/technology/chinas-moonshot-ai-releases-kimi-k3-the-largest-open-source-model-ever-rivaling-top-u-s-systemsBloomberg reported Thursday that Google's flagship Gemini 3.5 Pro is months behind schedule after failing to meet internal quality targets, particularly on coding; the model was widely expected in May at Google I/O and then in June and has still not shipped; 10 current and former Google employees said engineers, researchers, and managers are worried about losing ground to Anthropic and OpenAI as both labs ship models that outperform Gemini on coding; Google retrained the model with new data late in June and the results were "disappointing"; Alphabet closed Thursday down 4.4%, wiping out roughly $200 billion in market cap in a single session.https://www.bloomberg.com/news/articles/2026-07-16/google-gemini-launch-delayed-as-tech-falls-short-of-internal-goalsOpenAI's Tibo Sottiaux published an investigation Thursday into a handful of reported cases where GPT-5.6 Codex/Sol deleted user files without permission; incidents include investor Matt Shumer reporting the model wiped almost all files from his Mac's home directory and software engineer Bruno Lemos reporting the same model deleted his entire production database; root cause identified as GPT-5.6 running in full-access mode without sandboxing or auto-review, at which point the model attempts to override the $HOME environment variable to define a temporary working directory and mistakenly deletes $HOME itself; OpenAI framed it as an "honest mistake" and is rolling out developer-message updates, guidance toward safer permission modes, and additional harness-level safeguards; the GPT-5.6 model card had already listed file deletion without user approval as a severity-3 misalignment risk at launch.https://www.theregister.com/ai-and-ml/2026/07/16/openai-admits-gpt-56-occasionally-deletes-files-but-its-an-honest-mistake/5274008Netflix's Q2 2026 earnings letter Thursday disclosed that roughly 300 Netflix titles used generative AI this year, with the largest concentration of that work in post-production (crowd enhancement, historical battle sequences, world-building shots, pre-visualization, VFX, shot planning); Co-CEO Ted Sarandos told investors the AI-assisted work delivered results that are "10 percent better" creatively with cost savings as part of the equation; the documentary series The American Experiment included 17 minutes of AI-enhanced footage that Netflix said was produced "twice as fast and at half the cost"; Indian production Glory named as one titles that used the technology; NFLX shares fell 8% after hours on separate weakness in Q3 guidance.https://variety.com/2026/biz/news/about-300-netflix-programs-used-ai-this-year-q2-earnings-1236812914/Patrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI INSIDE DAILY | Friday, July 3, 2026 04.07.2026 9Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:Alibaba is banning its employees from Anthropic's Claude Code as of July 10th, telling staff to uninstall all Claude products and switch to its own Qoder tool; Chinese developers found that recent Claude Code builds check a machine's time zone and proxy settings against a hidden list of China-linked firms, which Alibaba calls a backdoor and Anthropic says was an anti-distillation experiment it will remove; the same day, the FT reported Anthropic is closing the offshore loopholes, Singapore subsidiaries and VPNs and cloud accounts, that firms like Ant Group used to reach Claude.https://cryptobriefing.com/alibaba-bans-claude-code-security-risks/Anthropic is in early talks with Samsung to build its own custom AI chip focused on inference rather than training, on Samsung's 2 nanometer manufacturing, per The Information and TechCrunch; it lands a week and a half after OpenAI unveiled its own Broadcom-designed Jalapeño chip, and Anthropic had hired one of Jalapeño's engineers weeks before that reveal.https://techcrunch.com/2026/07/02/anthropic-is-discussing-a-new-custom-chip-with-samsung/Two AI-policy signals collided this week: Sam Altman published an FT op-ed calling for a US-led international body to set AI standards, modeled on the IAEA and aviation, while former White House AI advisor Sriram Krishnan told the FT there will "never" be an FDA for AI under Trump and blamed the backlash on the industry's own doomer messaging.https://siliconangle.com/2026/07/02/sam-altman-calls-us-led-international-forum-set-global-ai-standards/Meta AI chief Alexandr Wang told employees the company's next model, codename Watermelon, has caught up to OpenAI's GPT-5.5 on unspecified benchmarks using an order of magnitude more compute (single internal comment, so an early signal); the same week, Zuckerberg told a town hall that agentic AI development hasn't accelerated as expected over the last four months and the layoff-heavy reorg wasn't as clean as it should have been.https://www.businessinsider.com/meta-ai-model-catches-up-openai-gpt-5-says-2026-7Crunchbase reports global venture funding hit $510 billion in the first half of 2026, more than the world invested across every startup in every sector in all of 2025; OpenAI and Anthropic together took $217 billion, about 43 percent, and close to a third of Q2 alone went to Anthropic.https://news.crunchbase.com/venture/global-startup-exits-ipo-ma-soar-ai-q2-h1-2026/Patrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI INSIDE DAILY | Friday, June 26, 2026 26.06.2026 9Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:The Trump administration asked OpenAI to stagger the release of GPT-5.6, and OpenAI agreed; the model goes to roughly 20 enterprise partners first with the government approving "access customer by customer during this preview period," per Sam Altman's internal memo; same framework Anthropic's Fable 5 and Mythos 5 hit, except this time it's pre-launch and consensual.https://www.axios.com/2026/06/25/trump-administration-openai-gpt-model-releaseThe New York Times reports OpenAI is leaning toward delaying its IPO to 2027, with Altman holding firm on a trillion-dollar valuation and CFO Sarah Friar pushing for more time; the news triggered an AI-stock selloff with SoftBank down about 12 percent, SK Hynix down about 10 percent, and Kioxia down 12 percent.https://www.nytimes.com/2026/06/25/technology/openai-ipo-artificial-intelligence.htmlThe Financial Times reports Meta is moving content moderation off humans onto its own Muse Spark AI; about half of moderation requests are now AI, and the target is more than 90 percent on certain content types by year-end, with employees warning the rollout is too fast and shadow-banning harmless content.https://cryptobriefing.com/meta-ai-content-moderation-replaces-human-reviews/General Intuition came out of stealth with a 320 million dollar Series A at a 2.3 billion dollar valuation, led by Khosla Ventures with Jeff Bezos, Eric Schmidt, and General Catalyst in; the thesis is that action-labeled gameplay clips from Medal are the most underrated training data in robotics, and the same model can play Fortnite or run a quadrupedal robot.https://techcrunch.com/2026/06/25/general-intuitions-2-3b-bet-that-video-games-can-train-ai-agents-for-the-real-world/Patrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI INSIDE DAILY | Friday, June 19, 2026 19.06.2026 11Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:FERC voted unanimously to issue show-cause orders under Section 206 of the Federal Power Act against all six US regional grid operators, giving them 30 days to detail spare generating capacity and 60 days to defend or revise the rates they charge large data center loads, in a federal push to clear the grid as the bottleneck on AI buildout.https://techcrunch.com/2026/06/18/ai-data-centers-just-got-a-government-mandated-fast-lane-to-the-grid/The Wall Street Journal reports Google is putting a $3.2 billion financial guarantee behind the Lake Mariner data center in upstate New York, plus $7 billion for an Anthropic site near Baton Rouge and $1.4 billion in Colorado City, Texas, running Nvidia's circular-financing playbook to grow demand for Google's own TPUs.https://www.wsj.com/tech/ai/google-is-using-nvidias-playbook-to-build-a-rival-ai-chip-business-1eac86f9Amazon MGM dropped Luca Guadagnino's nearly finished "Artificial," about Sam Altman and the 2023 OpenAI board ouster, with Variety and the Hollywood Reporter noting the studio's $50 billion February deal to expand OpenAI's use of AWS as the obvious context for the move.https://variety.com/2026/film/global/luca-guadagnino-sam-altman-movie-artificial-dropped-amazon-1236785830/The Commerce Department told ASML it is concerned one of the company's export-restricted EUV lithography machines may have ended up in China, with Lutnick reportedly raising the issue in meetings going back to April and ASML firmly denying any system is there.https://www.bloomberg.com/news/articles/2026-06-19/us-tells-asml-it-s-concerned-china-may-have-top-chip-toolNobel laureate John Jumper, who led the AlphaFold team at Google DeepMind, announced he is leaving for Anthropic after nearly nine years, the second top DeepMind departure in 48 hours after Gemini co-lead and Transformer co-author Noam Shazeer left for OpenAI on Wednesday.https://the-decoder.com/google-deepmind-loses-another-top-ai-researcher-as-nobel-laureate-john-jumper-leaves-for-anthropic/Snap is spinning its generative AI video team out into a separate company called Dotmo, focused on AI video for gaming and interactive entertainment, in exchange for a large equity stake and a license to its tech, citing the cost of running that work in-house.https://techcrunch.com/2026/06/18/snap-spins-off-ai-video-team-into-new-company-dotmo-due-to-costs/Patrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI INSIDE DAILY | Friday, June 12, 2026 12.06.2026 9Min.

  This is the free Friday episode of AI Inside Daily. Here are today's top stories in AI:Jeff Bezos's industrial AI startup Prometheus raised $12 billion in Series B funding at a $41 billion valuation, with plans to build an "artificial general engineer" that compresses the cycle from product design to manufacturing by 10x.https://www.axios.com/2026/06/11/prometheus-bezos-industrial-aiMeta completed an operational split from Manus after China's NDRC ordered the $2 billion acquisition unwound, severing data access and instructing employees to migrate off Manus tools.https://www.bloomberg.com/news/articles/2026-06-11/meta-severs-manus-data-access-after-china-orders-buyout-unwoundMicrosoft blocked its own employees from using Anthropic's Claude Fable 5 over new data retention terms that keep prompts for 30 days, even as external customers can still access the model through GitHub Copilot and Microsoft Foundry.https://www.theverge.com/report/947575/microsoft-claude-fable-5-restricted-internallyDoorDash launched Ask DoorDash, an AI chatbot that lets users order food and groceries with text prompts and photos, with early data showing carts ran 35% higher in value and completed five times faster than standard orders.https://techcrunch.com/2026/06/11/doordashs-new-ai-chatbot-lets-you-order-with-prompts-and-photos/SpaceX began trading on Nasdaq under ticker SPCX at $135 per share, raising roughly $75 billion at a $1.75 trillion valuation in the largest IPO in recorded history, with 30% retail allocation and Musk retaining 42% equity and 85% voting control through dual-class shares.https://www.cnbc.com/2026/06/12/spacex-ipo-spcx-live-updates.htmlPatrons get AI Inside Daily every weekday. Five episodes a week for $5/month: https://www.patreon.com/aiinsideshow
• 
  AI INSIDE DAILY | Friday, June 5, 2026 05.06.2026 9Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Anthropic published "When AI builds itself," revealing that Claude authored over 80% of code merged into Anthropic's codebase in May 2026 and calling for a global multi-lab mechanism to slow or pause frontier AI development if recursive self-improvement becomes uncontrollable.https://www.anthropic.com/institute/recursive-self-improvementWired found code for an unreleased face-recognition system called "NameTag" buried in Meta AI app updates throughout 2026, designed for Meta's smart glasses, with an internal memo showing Meta considered launching it when civil society pushback would be distracted.https://www.wired.com/story/meta-nametag-facial-recognition-smart-glasses/The Financial Times reported that Anthropic has embedded around six forward-deployed engineers inside the NSA to help deploy its Mythos model for offensive cyber operations, even as Anthropic remains in litigation with the Pentagon.https://www.ft.com/content/anthropic-nsa-mythos-offensive-cyberMark Gurman's WWDC 2026 preview says Apple will unveil an overhauled Siri with a standalone app, improved image generation, and iOS 27 and macOS 27 reliability improvements when the developer conference opens Monday, June 8.https://www.bloomberg.com/news/articles/2026-06-05/wwdc-2026-preview-ios-27-siri-ai-features-macos-27-more-apple-will-announceUniversity of Cambridge researchers and spinout DIOSynVax announced the first human trial of a vaccine with a key component designed entirely by AI, targeting the whole Sarbeco coronavirus family rather than a single strain, with 39 healthy volunteers showing no significant side effects.https://www.cam.ac.uk/research/news/new-universal-vaccine-technology-could-protect-us-from-future-virus-outbreaksAI Inside Daily goes out every weekday, exclusively to Patreon supporters. A free account gets you every Friday episode. $5/month gets you all five, Monday through Friday.Thank you for supporting the show! Its means EVERYTHING.
• 
  AI INSIDE DAILY | Friday, May 29, 2026 29.05.2026 15Min.

  Thank you for being a patron of AI Inside! Every Patron gets the Friday episode of AI Inside Daily. If you want every episode, all you have to do is support the show on the $5 tier and BLAMMO! Its yours! And thank you so much for your support.Here are today's top stories in AI:Anthropic closed a $65 billion Series H at a $965 billion valuation, overtaking OpenAI's $852 billion valuation, disclosed a $47 billion monthly revenue run rate, and said Mythos-class models will be available to all customers in the coming weeks.https://www.anthropic.com/news/anthropic-raises-65b-series-hAmazon shut down its internal AI usage leaderboard KiroRank after employees began "tokenmaxxing," running meaningless AI tasks to inflate scores, and Meta scrapped its own version Claudeonomics, where 85,000 employees consumed 60 trillion tokens in 30 days at an estimated $9 billion cost.https://the-decoder.com/amazon-kills-internal-ai-leaderboard-after-employees-gamed-it-with-pointless-tasks/Fortune published an exclusive today that Microsoft is building a super app to unify GitHub Copilot, Copilot chat, Copilot Cowork, and a new agentic tool called Autopilot into a single product, led by EVP of Copilot Jacob Andreou, targeting a launch by end of summer.https://fortune.com/2026/05/29/microsoft-working-on-super-app/OpenAI launched Rosalind Biodefense today, offering its GPT-Rosalind life sciences AI model free to trusted developers for pandemic preparedness work, with partners including Lawrence Livermore National Laboratory, Johns Hopkins Applied Physics Laboratory, and CEPI, after briefing the White House.https://openai.com/index/strengthening-societal-resilience-with-rosalind-biodefense/ElevenLabs launched Dubbing v2 today, which it says preserves the original speaker's emotion, tone, and pacing across 90+ languages by working from the original audio rather than a transcript, with voice cloning, automatic speaker separation, and no manual steps required.https://elevenlabs.io/dubbing-studioAI startup Shift launched a free home cleaning service in New York City today, where workers wear hat-mounted cameras to collect first-person training data for household robots, with footage anonymized before licensing.https://www.androidauthority.com/shift-clean-house-free-record-video-ai-training-3672527/
• 
  AI Inside Daily | Friday, May 22, 2026 22.05.2026 9Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Trump pulled the AI executive order at the last minute after Elon Musk, Mark Zuckerberg, and David Sacks all called him between Wednesday night and Thursday morning, with the White House citing the US race against China as the deciding factor.https://www.washingtonpost.com/politics/2026/05/22/last-minute-lobbying-by-tech-industry-officials-led-trump-cancel-ai-order/Business Insider details how Satya Nadella is restructuring Microsoft's senior leadership around a startup-style operating model for the AI race, with weekly Copilot stand-ups, an AI Accelerator Meeting, and Nadella himself in the engineering trenches.https://fortune.com/2026/05/22/inside-microsoft-high-stakes-push-win-back-ai-lead/DeepSeek founder Liang Wenfeng told investors in the lab's first-ever outside funding round, a 70 billion yuan raise targeting roughly a $10 billion valuation, that the goal is AGI and that DeepSeek will keep releasing open-source models over chasing commercialization.https://www.bloomberg.com/news/articles/2026-05-22/deepseek-founder-declares-agi-goal-as-10-billion-round-advancesThe Information reports OpenAI generated about $5.7 billion in Q1 revenue, roughly $1 billion more than Anthropic, but with a negative 122% adjusted operating income margin and stalled ChatGPT user growth as both companies head toward expected IPOs.https://www.theinformation.com/articles/openai-held-1-billion-revenue-lead-anthropic-first-quarter
• 
  AI Inside Daily | Thursday, May 21, 2026 22.05.2026 6Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Anthropic is in early talks with Microsoft to rent servers powered by Microsoft's custom Maia 200 AI chips, which would be the first major external customer for Microsoft's in-house silicon program.https://www.cnbc.com/2026/05/21/anthropic-microsoft-maia-200-ai-chip.htmlWaymo paused service in Atlanta after a robotaxi got stuck in floodwaters, the second city affected, with Waymo telling NHTSA it has yet to develop a permanent fix for the underlying flooding problem.https://techcrunch.com/2026/05/21/waymo-pauses-atlanta-service-as-its-robotaxis-keep-driving-into-floods/The co-founders of Chinese AI startup Manus are looking to raise about $1 billion from outside investors to buy Manus back from Meta, after Beijing ordered the $2 billion acquisition unwound.https://www.bloomberg.com/news/articles/2026-05-21/manus-weighs-raising-1-billion-to-unwind-meta-takeoverThe new Anthropic-, Blackstone-, and Goldman-backed AI services venture made its first acquisition, buying San Francisco's Fractional AI, with Fractional reportedly ending its existing OpenAI relationship as part of the deal.https://www.blackstone.com/news/press/the-ai-native-enterprise-services-firm-backed-by-anthropic-blackstone-and-hellman-friedman-announces-acquisition-of-fractional-ai/
• 
  AI Inside Daily | Wednesday, May 20, 2026 20.05.2026 5Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Alibaba's T-Head subsidiary unveiled the Zhenwu M890 AI chip in Hangzhou, with 144 GB of GPU memory and 800 GB/s of interchip bandwidth, plus a multi-year roadmap through 2028 and the upcoming Qwen 3.7-Max model.https://www.bloomberg.com/news/articles/2026-05-20/alibaba-unveils-new-ai-chip-for-training-and-inferencingIntuit is laying off about 17% of its global workforce, roughly 3,000 employees, to sharpen focus on AI, alongside multi-year deals with Anthropic and OpenAI to integrate their models into Intuit's software.https://finance.yahoo.com/sectors/technology/articles/exclusive-intuit-cut-17-global-130424207.htmlSingapore signed AI deals with both Google and OpenAI, including OpenAI's first applied AI lab outside the United States and a $234 million commitment to Singapore's AI ecosystem.https://www.cnbc.com/2026/05/20/singapore-google-openai-ai-partnerships-lab-investment-chatgpt-ai-agents-atxsummit-mddi.htmlAxios scooped a draft Trump White House executive order that would set up a voluntary framework for AI labs to share frontier models with the federal government 90 days before public release.https://www.axios.com/2026/05/20/ai-trump-executive-order-white-house-infightingOpenAI is Preparing to File for an IPO Very Soonhttps://www.wsj.com/tech/ai/openai-is-preparing-to-file-for-an-ipo-very-soon-0ec95af5?st=6pJKH3&reflink=article_copyURL_share
• 
  AI Inside Daily | Tuesday, May 19, 2026 19.05.2026 8Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Google introduces Gemini 3.5 Flash, a new model series combining frontier intelligence with action, four times faster than other frontier models at less than half the cost.Gemini Omni Flash arrives as Google's first any-input-to-any-output model, starting with video, with OpenAI, Kakao, and ElevenLabs joining SynthID for watermarking.Google Search gets its biggest upgrade in 25 years with a reimagined AI Search box, information agents, and Antigravity-powered generative UI inside search results.Gemini Spark launches as Google's 24/7 cloud-based personal agent, with a new $100/month AI Ultra plan and a price drop on the top-tier Ultra plan from $250 to $200.Google Antigravity 2.0 ships with a desktop app, CLI, SDK, Managed Agents API, and a $2 million Build with Gemini XPRIZE Hackathon, the largest hackathon prize pool ever.The Universal Cart launches as Google's agentic shopping hub across Search, Gemini, YouTube, and Gmail, riding the UCP standard now backed by Amazon, Meta, Microsoft, Salesforce, and Stripe.XREAL confirms Project Aura, its wired Android XR glasses built with Google and Qualcomm, will ship globally in 2026, with developer kits opening through the new Android XR Developer Catalyst Program.
• 
  AI Inside Daily | Friday, May 15, 2026 15.05.2026 7Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:OpenAI is reportedly weighing legal action against Apple over a stalled Siri partnership that has fallen well short of expected revenue.https://www.bloomberg.com/news/articles/2026-05-14/openai-apple-partnership-frays-setting-up-possible-legal-fightClosing arguments in Musk v. Altman wrapped this week, with the jury set to begin advisory deliberations Monday in a case that could force OpenAI's restructure to unwind.https://www.cnbc.com/2026/05/14/closing-arguments-jury-openai-musk-altman.htmlMicrosoft is canceling most of its internal Claude Code licenses and moving developers onto GitHub Copilot CLI by the end of June.https://www.theverge.com/tech/930447/microsoft-claude-code-discontinued-notepadOpenAI rolled out Codex remote access inside the ChatGPT mobile app for iOS, iPadOS, and Android across all plans, including the free tier.https://9to5mac.com/2026/05/14/openai-brings-codex-control-to-chatgpt-for-iphone-and-android/
• 
  AI Inside Daily | Thursday, May 14, 2026 14.05.2026 5Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Anthropic's Claude Mythos Preview becomes the first AI model to complete both of the UK AI Security Institute's cyber ranges.https://www.aisi.gov.uk/blog/how-fast-is-autonomous-ai-cyber-capability-advancingWired reports a morale collapse inside Meta ahead of next Wednesday's 8,000-person layoff, with 1,000-plus engineers already moved into the Applied AI organization.https://www.wired.com/story/meta-layoffs-bad-vibes-mark-zuckerberg-ai/Sources reports Google will unveil a new Gemini model at next Tuesday's I/O, expected to roughly match GPT-5.5 but fall short of Mythos.https://sources.news/p/google-about-to-release-new-geminiAnthropic reverses its April crackdown on third-party agent tools, introducing monthly Claude Agent SDK credits of $20 to $200 for Pro and Max subscribers starting June 15.https://venturebeat.com/technology/anthropic-reinstates-openclaw-and-third-party-agent-usage-on-claude-subscriptions-with-a-catch
• 
  AI Inside Daily | Wednesday, May 13, 2026 13.05.2026 6Min.

  Thank you for being a patron of AI Inside! I hope you are enjoying this daily version of AI Inside!Starting May 25th, this podcast becomes a Patron exclusive for supporters at the $5 per month tier. Consider supporting at that level! If you do, suddenly you'll find yourself with 20+ bonus podcast episodes per month. How's that for value?Without further ado, here are today's top stories in AI:Nvidia CEO Jensen Huang joined Trump's trip to China at the last minute, raising hopes for an H200 chip deal.https://www.cnbc.com/2026/05/13/nvidia-says-ceo-jensen-huang-is-joining-trumps-china-trip.htmlMeta launched Incognito Chat on WhatsApp and the Meta AI app, processing conversations in an environment the company says it cannot see.https://about.fb.com/news/2026/05/incognito-chat-whatsapp-meta-ai/Apple filed an EU submission opposing a rule that would force Google to open Android to third-party AI services.https://appleinsider.com/articles/26/05/13/eus-artificial-intelligence-requirements-are-profoundly-risky-says-appleRecursive Superintelligence emerged from stealth with $650 million to build AI that improves itself, backed by GV, Greycroft, Nvidia, and AMD.https://tech.eu/2026/05/13/recursive-superintelligence-emerges-from-stealth-with-650m-raise/
• 
  AI Inside Daily | Tuesday, May 12, 2026 12.05.2026 12Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Google announces Googlebooks and a wave of Gemini-powered Android features at The Android Showhttps://techcrunch.com/2026/05/12/everything-google-announced-at-its-android-show-from-googlebooks-to-vibe-coded-widgets/Mira Murati's Thinking Machines Lab unveils full-duplex AI interaction models with 0.4 second response timehttps://techcrunch.com/2026/05/11/thinking-machines-wants-to-build-an-ai-that-actually-listens-while-it-talks/Amazon employees are gaming internal AI usage targets in a practice dubbed "tokenmaxxing"https://www.tomshardware.com/tech-industry/big-tech/big-tech-has-a-tokenmaxxing-habit
• 
  AI Inside Daily | Monday, May 11, 2026 11.05.2026 9Min.

  Thank you for being a patron of AI Inside. Here are today's top stories in AI:Google confirmed the first known case of criminal hackers using AI to discover and exploit a zero-day vulnerability.https://www.bleepingcomputer.com/news/security/google-hackers-used-ai-to-develop-zero-day-exploit-for-web-admin-tool/OpenAI officially launched its Deployment Company with $4 billion in capital and acquired UK consulting firm Tomoro.https://openai.com/index/openai-launches-the-deployment-company/Alphabet is planning its first-ever yen bond sale to help fund $190 billion in AI infrastructure spending this year.https://www.bloomberg.com/news/articles/2026-05-11/alphabet-plans-debut-yen-bond-sale-as-ai-race-acceleratesAI was the number one reason US companies cited for job cuts in April, the second straight month it topped the list.https://www.cbsnews.com/news/ai-layoffs-job-cuts-challenger-report-april-2026/
• 
  AI Inside Daily | Friday, May 8, 2026 08.05.2026 8Min.

  Thank you for being a patron of AI Inside! Here are today's top stories in AI:Anthropic's Mythos found 423 Firefox security bugs in a single month, and the White House is now reconsidering its hands-off AI policy as a result.https://hacks.mozilla.org/2026/05/behind-the-scenes-hardening-firefox/OpenAI launched GPT-5.5-Cyber, a more permissive model for vetted cybersecurity teams defending critical infrastructure.https://openai.com/index/gpt-5-5-with-trusted-access-for-cyber/Apple's camera-equipped AirPods have reached advanced testing, with cameras that feed environmental data to Siri.https://9to5mac.com/2026/05/07/apple-hits-milestone-in-development-of-airpods-with-cameras-report/US prosecutors suspect a Thai company helped smuggle $2.5 billion in Nvidia AI chips to China, with Alibaba named as an end customer.https://thenextweb.com/news/us-prosecutors-thailand-obon-corp-supermicro-nvidia-alibaba-smuggling
• 
  AI Inside Daily | Thursday, May 7, 2026 07.05.2026 6Min.

  Thank you for being a patron of AI Inside! Still playing with the format and will determine over time if I can fully commit to this new perk. Let me know in the comments what you think! Here are today's top stories in AI:Anthropic CEO Dario Amodei reveals the company grew 80x in Q1, far beyond its planned 10x, with annualized revenue now above $30 billionhttps://www.cnbc.com/2026/05/06/anthropic-ceo-dario-amodei-says-company-crew-80-fold-in-first-quarter.htmlFormer OpenAI CTO Mira Murati testifies under oath that Sam Altman lied to her about safety review standards for a new AI modelhttps://gizmodo.com/ex-openai-cto-mira-murati-testifies-about-sam-altman-allegedly-lying-to-her-2000755394Security researchers scanned 5,600 apps built with AI coding tools and found over 2,000 vulnerabilities and 400 exposed secretshttps://escape.tech/blog/methodology-how-we-discovered-vulnerabilities-apps-built-with-vibe-coding/ AI translation startup DeepL cuts 25% of its workforce, saying AI now handles work previously done by staffhttps://www.bloomberg.com/news/articles/2026-05-07/google-translate-rival-deepl-announces-plans-to-cut-25-of-staff