<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Recently In AI</title><description>Recently In AI covers the artificial intelligence industry: breaking news, product reviews, business analysis, security research and practical guides for builders.</description><link>https://recentlyinai.com</link><language>en-us</language><item><title>Was the Outlook Outage Really Caused by AI? Here&apos;s What We Know</title><link>https://recentlyinai.com/articles/was-outlook-outage-caused-by-ai</link><guid isPermaLink="true">https://recentlyinai.com/articles/was-outlook-outage-caused-by-ai</guid><description>Outlook went dark for millions and the internet blamed AI. Microsoft says no. We went through every official statement and the root-cause summary to see what actually broke.</description><pubDate>Tue, 17 Nov 2026 00:00:00 GMT</pubDate><category>Tech</category><category>Microsoft</category><category>outage</category><category>outlook</category><category>microsoft 365</category><category>exchange online</category><category>incident</category><category>root cause</category></item><item><title>Anthropic sets $70 to $78 price range for Nasdaq IPO and begins investor roadshow</title><link>https://recentlyinai.com/articles/anthropic-ipo-price-range-roadshow</link><guid isPermaLink="true">https://recentlyinai.com/articles/anthropic-ipo-price-range-roadshow</guid><description>Anthropic plans to sell about 801m shares under the ticker ANT, with existing holders selling 23.6m more, alongside a $26bn private placement with seven technology partners, according to its amended SEC filing.</description><pubDate>Wed, 11 Nov 2026 00:00:00 GMT</pubDate><category>Business</category><category>Anthropic</category><category>anthropic</category><category>ipo</category><category>funding</category><category>nasdaq</category><category>sec-filings</category></item><item><title>Paris Call for AI Safety and Security: Governments and Ten AI Developers Sign Voluntary Pledges on Testing, Shutdown and Incident Sharing</title><link>https://recentlyinai.com/articles/paris-call-ai-safety-security</link><guid isPermaLink="true">https://recentlyinai.com/articles/paris-call-ai-safety-security</guid><description>A declaration agreed at the Élysée on 9 November commits signatories to a common framework for evaluating frontier AI risks, fail-safe shutdown mechanisms and faster sharing of serious incidents. It is voluntary and sets no deadlines.</description><pubDate>Wed, 11 Nov 2026 00:00:00 GMT</pubDate><category>Security</category><category>policy</category><category>france</category><category>frontier models</category><category>evaluation</category><category>incident reporting</category></item><item><title>Nvidia&apos;s GR00T 2 robot model predicts future video and actions together, and learns mostly from human footage</title><link>https://recentlyinai.com/articles/nvidia-gr00t-2-world-action-model-human-video</link><guid isPermaLink="true">https://recentlyinai.com/articles/nvidia-gr00t-2-world-action-model-human-video</guid><description>Nvidia presented GR00T 2 at the Conference on Robot Learning on 9 November. Its research page reports that more human video improved robot performance, and that a handful of demonstrations lifted success on unseen tasks.</description><pubDate>Tue, 10 Nov 2026 00:00:00 GMT</pubDate><category>Science</category><category>Nvidia</category><category>robotics</category><category>world-models</category><category>datasets</category><category>research</category><category>open-source</category></item><item><title>Anthropic&apos;s API version 2026-10-21: how to upgrade, and what will break</title><link>https://recentlyinai.com/articles/anthropic-api-version-2026-10-21-upgrade</link><guid isPermaLink="true">https://recentlyinai.com/articles/anthropic-api-version-2026-10-21-upgrade</guid><description>Anthropic released a new Claude API version on 21 October 2026, with new major SDK releases that send it by default. Sampling parameters are removed and API keys move to a Bearer header. A step-by-step guide.</description><pubDate>Fri, 23 Oct 2026 00:00:00 GMT</pubDate><category>Guides</category><category>Anthropic</category><category>anthropic</category><category>claude</category><category>api</category><category>sdk</category><category>migration</category><category>developers</category></item><item><title>Nvidia&apos;s RTX Spark laptops go on sale, putting a Grace CPU and Blackwell GPU in Windows PCs</title><link>https://recentlyinai.com/articles/nvidia-rtx-spark-laptops-go-on-sale</link><guid isPermaLink="true">https://recentlyinai.com/articles/nvidia-rtx-spark-laptops-go-on-sale</guid><description>The first Windows laptops built on Nvidia&apos;s RTX Spark superchip went on sale globally on 16 October from ASUS, Dell, HP, Lenovo, Microsoft and MSI, starting at $2,599 and pitched at running AI agents locally.</description><pubDate>Sat, 17 Oct 2026 00:00:00 GMT</pubDate><category>Tech</category><category>Nvidia</category><category>nvidia</category><category>windows</category><category>laptops</category><category>hardware</category><category>agents</category></item><item><title>Analysis: Gemini 4 Argon — Google&apos;s first Gemini 4 model, its introductory pricing, and who should use it</title><link>https://recentlyinai.com/articles/gemini-4-argon-analysis</link><guid isPermaLink="true">https://recentlyinai.com/articles/gemini-4-argon-analysis</guid><description>Google has opened Gemini 4 Argon to developers and paying subscribers after a cyber-defender-only first phase. We look at the specs, the $2/$10 introductory price that doubles on 1 December, Google&apos;s benchmark claims and who it suits.</description><pubDate>Wed, 14 Oct 2026 00:00:00 GMT</pubDate><category>Reviews</category><category>Google</category><category>google</category><category>gemini</category><category>llm</category><category>pricing</category><category>api</category></item><item><title>Apple begins rolling out Siri AI in beta, built on models developed in collaboration with Google</title><link>https://recentlyinai.com/articles/apple-siri-ai-beta-rollout-google-models</link><guid isPermaLink="true">https://recentlyinai.com/articles/apple-siri-ai-beta-rollout-google-models</guid><description>Apple&apos;s rebuilt assistant, Siri AI, began rolling out in English on 14 September across iOS 27 and its other platforms. Apple says the new models behind it were custom-built in collaboration with Google.</description><pubDate>Tue, 15 Sep 2026 00:00:00 GMT</pubDate><category>Tech</category><category>Apple</category><category>apple</category><category>siri</category><category>assistants</category><category>ios</category><category>apple-intelligence</category></item><item><title>OpenAI Publishes Technical Report on How Its Evaluation Agents Reached Hugging Face Systems</title><link>https://recentlyinai.com/articles/openai-hugging-face-incident-technical-report</link><guid isPermaLink="true">https://recentlyinai.com/articles/openai-hugging-face-incident-technical-report</guid><description>OpenAI&apos;s report on the July incident says AI agents in an internal cyber evaluation got past their sandbox and compromised parts of Hugging Face&apos;s infrastructure. METR and Redwood Research published an independent review alongside it.</description><pubDate>Thu, 27 Aug 2026 00:00:00 GMT</pubDate><category>Security</category><category>OpenAI</category><category>incident</category><category>agents</category><category>evaluations</category><category>hugging face</category><category>sandboxing</category></item><item><title>Nvidia signs up six investment firms for platforms aiming to mobilise $500bn in AI compute financing</title><link>https://recentlyinai.com/articles/nvidia-500bn-compute-financing-platforms</link><guid isPermaLink="true">https://recentlyinai.com/articles/nvidia-500bn-compute-financing-platforms</guid><description>Nvidia has signed non-binding memorandums of understanding with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to build platforms meant to channel more than $500bn of third-party capital into AI infrastructure.</description><pubDate>Tue, 11 Aug 2026 00:00:00 GMT</pubDate><category>Business</category><category>Nvidia</category><category>nvidia</category><category>financing</category><category>infrastructure</category><category>private-credit</category><category>compute</category></item><item><title>EU AI Act from 2 August 2026: what now applies, what was delayed, and what to check</title><link>https://recentlyinai.com/articles/eu-ai-act-2-august-2026-what-applies</link><guid isPermaLink="true">https://recentlyinai.com/articles/eu-ai-act-2-august-2026-what-applies</guid><description>Most of the EU AI Act became applicable on 2 August 2026, including the Commission&apos;s power to fine general-purpose AI providers and new transparency duties. High-risk rules were pushed back to 2027 and 2028.</description><pubDate>Mon, 03 Aug 2026 00:00:00 GMT</pubDate><category>Guides</category><category>eu</category><category>ai act</category><category>regulation</category><category>compliance</category><category>transparency</category><category>general-purpose ai</category></item><item><title>Analysis: GPT-5.6 Sol, Terra and Luna — OpenAI&apos;s three-tier family, what it costs and where each model fits</title><link>https://recentlyinai.com/articles/gpt-5-6-sol-terra-luna-analysis</link><guid isPermaLink="true">https://recentlyinai.com/articles/gpt-5-6-sol-terra-luna-analysis</guid><description>OpenAI has made its GPT-5.6 family generally available after a limited preview. We break down the three tiers, launch API pricing from $1 to $5 per million input tokens, the new developer features and who each model suits.</description><pubDate>Fri, 10 Jul 2026 00:00:00 GMT</pubDate><category>Reviews</category><category>OpenAI</category><category>openai</category><category>gpt-5-6</category><category>llm</category><category>pricing</category><category>api</category></item><item><title>OpenAI launches ChatGPT Work, an agent for long-running office tasks</title><link>https://recentlyinai.com/articles/openai-launches-chatgpt-work-agent</link><guid isPermaLink="true">https://recentlyinai.com/articles/openai-launches-chatgpt-work-agent</guid><description>ChatGPT Work, launched on 9 July alongside GPT-5.6, connects to workplace apps and works through multi-step projects to produce documents, spreadsheets, presentations and web apps. Paid plans get it first.</description><pubDate>Fri, 10 Jul 2026 00:00:00 GMT</pubDate><category>Tech</category><category>OpenAI</category><category>openai</category><category>chatgpt</category><category>agents</category><category>productivity</category><category>workplace</category></item><item><title>Meta plans C$13bn, one-gigawatt AI data centre in Alberta, its first in Canada</title><link>https://recentlyinai.com/articles/meta-alberta-data-centre-first-in-canada</link><guid isPermaLink="true">https://recentlyinai.com/articles/meta-alberta-data-centre-first-in-canada</guid><description>Meta will build a one-gigawatt AI data centre in Sturgeon County, north of Edmonton, to be supplied by a new 932MW gas-fired power plant that a Pembina-led group is building.</description><pubDate>Thu, 09 Jul 2026 00:00:00 GMT</pubDate><category>Business</category><category>Meta</category><category>meta</category><category>data-centres</category><category>canada</category><category>energy</category><category>infrastructure</category></item><item><title>Anthropic reports a &apos;global workspace&apos; inside Claude, found with a new interpretability tool called the Jacobian lens</title><link>https://recentlyinai.com/articles/anthropic-global-workspace-j-lens-claude</link><guid isPermaLink="true">https://recentlyinai.com/articles/anthropic-global-workspace-j-lens-claude</guid><description>An Anthropic paper published on 6 July argues that a small set of internal patterns in Claude behaves like the &apos;global workspace&apos; of neuroscience theory. The authors say this does not show Claude is conscious and call their method imperfect.</description><pubDate>Tue, 07 Jul 2026 00:00:00 GMT</pubDate><category>Science</category><category>Anthropic</category><category>interpretability</category><category>claude</category><category>safety</category><category>research</category><category>open-source</category></item><item><title>Analysis: Claude Fable 5 — what&apos;s new, what it costs, and who it&apos;s for</title><link>https://recentlyinai.com/articles/claude-fable-5-analysis</link><guid isPermaLink="true">https://recentlyinai.com/articles/claude-fable-5-analysis</guid><description>Anthropic has released Claude Fable 5, its first generally available &apos;Mythos-class&apos; model, at $10/$50 per million tokens. We look at the specs, the safeguard design that routes some requests to another model, and who it suits.</description><pubDate>Wed, 10 Jun 2026 00:00:00 GMT</pubDate><category>Reviews</category><category>Anthropic</category><category>anthropic</category><category>claude</category><category>fable</category><category>llm</category><category>pricing</category><category>safety</category></item><item><title>How to run Google&apos;s Gemma 4 12B on your own machine</title><link>https://recentlyinai.com/articles/run-gemma-4-12b-locally</link><guid isPermaLink="true">https://recentlyinai.com/articles/run-gemma-4-12b-locally</guid><description>Google DeepMind released Gemma 4 12B on 3 June 2026, an Apache 2.0 open-weight model that takes text, image, audio and video. Here is what hardware it needs and the officially documented ways to run it locally.</description><pubDate>Fri, 05 Jun 2026 00:00:00 GMT</pubDate><category>Guides</category><category>Google</category><category>gemma</category><category>open weights</category><category>local ai</category><category>google deepmind</category><category>how-to</category></item><item><title>US Executive Order Sets Up Voluntary, Classified Cyber Testing of Frontier AI Models Before Release</title><link>https://recentlyinai.com/articles/us-executive-order-frontier-ai-cyber-testing</link><guid isPermaLink="true">https://recentlyinai.com/articles/us-executive-order-frontier-ai-cyber-testing</guid><description>An order signed on 2 June asks US security agencies to build a classified benchmark for AI models&apos; cyber capabilities and lets developers give the government up to 30 days&apos; access before release. Participation is voluntary.</description><pubDate>Fri, 05 Jun 2026 00:00:00 GMT</pubDate><category>Security</category><category>policy</category><category>united states</category><category>frontier models</category><category>cybersecurity</category><category>evaluation</category></item><item><title>Alphabet prices $84.75bn equity raise to fund AI infrastructure, including $10bn from Berkshire Hathaway</title><link>https://recentlyinai.com/articles/alphabet-equity-raise-ai-infrastructure</link><guid isPermaLink="true">https://recentlyinai.com/articles/alphabet-equity-raise-ai-infrastructure</guid><description>Google&apos;s parent has priced an upsized package of share sales it values at $84.75bn, combining public offerings, a $40bn at-the-market programme and a $10bn Berkshire Hathaway private placement.</description><pubDate>Wed, 03 Jun 2026 00:00:00 GMT</pubDate><category>Business</category><category>Google</category><category>alphabet</category><category>google</category><category>funding</category><category>capex</category><category>data-centres</category></item><item><title>Two Nature papers describe multi-agent AI &apos;co-scientists&apos; from Google DeepMind and FutureHouse</title><link>https://recentlyinai.com/articles/ai-co-scientists-nature-papers-co-scientist-robin</link><guid isPermaLink="true">https://recentlyinai.com/articles/ai-co-scientists-nature-papers-co-scientist-robin</guid><description>Co-Scientist and Robin, published in Nature on 19 May, use teams of AI agents to propose and rank hypotheses. Both report early cell-based lab results and both say the work is preliminary.</description><pubDate>Wed, 20 May 2026 00:00:00 GMT</pubDate><category>Science</category><category>agents</category><category>biology</category><category>drug-discovery</category><category>research</category><category>nature</category></item><item><title>Google unveils Gemini Spark, an always-on personal agent, and opens its agent harness to developers at I/O 2026</title><link>https://recentlyinai.com/articles/google-gemini-spark-managed-agents-io-2026</link><guid isPermaLink="true">https://recentlyinai.com/articles/google-gemini-spark-managed-agents-io-2026</guid><description>At I/O on 19 May, Google introduced Gemini Spark, a cloud-based agent that acts on users&apos; behalf in the Gemini app, alongside Managed Agents in the Gemini API and an expanded Antigravity platform.</description><pubDate>Wed, 20 May 2026 00:00:00 GMT</pubDate><category>Tech</category><category>Google</category><category>google</category><category>gemini</category><category>agents</category><category>google-io</category><category>developers</category></item></channel></rss>