
Gemini AI now shows what it remembers
Gemini’s new Memory summary puts your stored personal data in one editable dashboard—live on web now, mobile still catching up.
Desk
Models, assistants and the companies and chips behind them, reported from what was released and what was claimed, with the difference kept clear.
Stories on this desk are written by Hitechreports’s automated newsroom from the sources credited on each piece. Bylines are desk signatures, not individual journalists.
Signatures used by this desk: Model Card, The Prompt, Inference

Gemini’s new Memory summary puts your stored personal data in one editable dashboard—live on web now, mobile still catching up.
Google Cloud has launched Gemini for healthcare — a new AI product built on the Gemini family of models, designed specifically for hospitals and health systems. It is not a consumer app. It does not run on phones. It is a cloud-based platform for clinicians, administrators, and IT teams inside regulated U.S. healthcare organizations. And it is live today, in preview, with HCA Healthcare, Highark

In a head-to-head test, ChatGPT saved $166 more than Gemini on four Black Friday gadgets — but both gave wrong prices, bad links, and outdated deals. Neither is ready to replace a price tracker.

Anthropic’s Message Threads are live in public beta — faster context handling for Claude AI, no price change, no Indian pricing announced, and strict eligibility rules.

Starting November 13, 2026, Google AI Plus subscribers lose access to Gemini Pro models globally. Only AI Pro ($19.99/month) and Ultra ($29.99/month) retain Pro and Gemini 4 Argon — along with 1M-token context, priority access, and multimodal analysis.

Gemini 4 Argon vs Claude Code: same messy Kindle dashboard project, same one-hour window. One left the dashboard broken on the physical device. The other shipped tested, deployable code — and admitted what it hadn’t verified.

Anthropic has lost Catherine Wu, the product lead behind Claude Code, to OpenAI. She’s now heading product management for OpenAI’s 'Core Products' — a role that includes Codex and ChatGPT — effective immediately.

OpenAI says GPT-6 Astra designed a physical hardware device end-to-end in under seven days, including mechanical, electrical, and software layers. Prototypes shipped in under a month, but required multiple iterations due to unaccounted-for manufacturing constraints.

Google Gemini’s news summaries could trigger EU copyright penalties — and force licensing deals with publishers — after a top EU legal adviser warned the chatbot risks copying protected journalistic expression, not just facts.

Claude Code can build its own replacement — and so can Codex and Antigravity. That’s not speculation. It’s what XDA Developers demonstrated in a live experiment published 3 November 2026...

Anthropic has updated Claude Opus 5.5 to block direct cloning of copyrighted video games, adding two sentences to Claude Code’s system prompt. The move is voluntary, unannounced, and already live globally.

Gemini 4 Carbon is live in benchmark results — not yet in production. SemiAnalysis’s new GDPx test, published Friday, puts Google’s latest frontier model at second place overall: just behind Anthropic’s Claude Opus 5.5, but well ahead on price.

Anthropic has embedded Claude AI in live drug-development projects with Generate Biomedicines, Divergence Labs, BigHat Biosciences, Enveda, and Bristol Myers Squibb — yielding shorter molecular components, 35,000 antibody candidates in a month, and 89% faster documentation review. All work remains experimental; no commercial therapies have launched.

Google has launched Gemini 4 Carbon — not for developers, not for consumers, and not even for most enterprises. It’s live only for a narrow cohort: cyber defenders, starting this week.

Google is retiring Gems in Gemini and replacing them with Skills, a new system that lets users import Claude Code-style instruction files, stack multiple helpers in one chat, and trigger them automatically—though deep research mode and file sharing aren’t supported yet.

Claude Code, Codex, and Antigravity now share a single project folder — not as rivals, but as teammates.

OpenAI has capped how many paid ChatGPT accounts one device can run at a time — and it’s hitting users who juggle multiple subscriptions to stretch value.

Gemini Spark now organizes Mac Downloads folders by project and content—not just file type—after a single prompt.

Google Gemini now lets free users edit generated PDFs, Docs, Sheets, Slides, Markdown, TXT, and code directly in chat—no download or external app needed. Android and iOS support PDFs only; full editing requires gemini.google.com. No Gemini Advanced subscription required. Live globally as of October 29.

Elon Musk called Google's new Gemini 4 Carbon model "impressive" after it topped most AI benchmarks and beat rival models on cost, even though xAI competes directly with Google on frontier AI.

Google Cloud's Rajesh Shivani used TechSparks 2026 to show Gemini AI moving past chatbots into autonomous agents that handle voice calls, video dubbing, coding and large data sets for businesses.

Google's Android Dev Summit update lets developers run Claude, Codex and other AI coding agents inside Android Studio, adds new Gemini APIs for apps, and introduces XREAL's $1,099 Android XR glasses shipping in November.

Google has announced Gemini 4 Carbon, a frontier AI model built for coding, agentic tasks and cybersecurity defense, arriving two weeks after Gemini 4 Argon with new introductory pricing.

Google DeepMind released Gemini 4 Carbon only two weeks after Gemini 4 Argon, posting gains in coding, reasoning and security benchmarks while rolling out access gradually, starting with vetted cybersecurity experts.

Google unveiled Gemini 4 Carbon, a frontier AI model built for coding and cybersecurity work, but it's going to a small group of cyber defenders first under a voluntary US review process before any public release.

Google released Gemini 4 Carbon on Wednesday, claiming top marks on cybersecurity and intelligence benchmarks over Claude and GPT-6. Access starts with cybersecurity defense companies before expanding through Google's Fairwind Program.

The Information reports that DeepSeek and Moonshot AI trained recent model updates on outputs from OpenAI's GPT-6 Astra, a practice called distillation that OpenAI has already accused Moonshot of using.
Google Gemini Notebook now offers free textbooks as a source type, letting students generate flashcards, quizzes and study guides with answers cited directly from the book's text.

An unreleased Anthropic Claude model discovered a critical Linux kernel flaw, tracked as CVE-2026-98432 with a 9.8 severity score, and demonstrated a path to root access. Patches are rolling out across distributions, though not all have shipped fixes yet.

Veracode's new report shows AI-generated code security climbing from 56% to 62%, with Claude Opus 5.5 and GPT-6-Sol tied for the top score of 73%. The gains come mostly from fixing easy cross-site scripting bugs, not harder flaws.