New features and launches in ChatGPT, Cursor, Perplexity and other AI tools, plus changes to pricing and availability.
Fambot just launched an "AI chief of staff" designed to tackle the mental load of parenting. Built by former Uber and Instagram veterans, the app hooks into your email, calendar, and WhatsApp groups to proactively generate daily checklists and track school activities. Unlike text-only agents, it offers full mobile and web interfaces and is currently free in beta.
OpenAI just rolled out Epic EHR integration and a Healthcare Public Data plugin for ChatGPT for Healthcare. Clinicians can now pull authorized lab results, medication changes, and appointment notes straight into the chat or inside Epic's chart layout, while tapping into nine official sources like PubMed. In physician reviews across 4,363 ratings, 99.1% of responses were deemed safe.
Dyson just revealed the CameraJet, an electric toothbrush equipped with a 100k-pixel macro lens and Gap Optical Targeting AI. It scans 28 frames per second to spot gaps between teeth in real time, firing a conical jet of mouthwash within 100ms. After six years of R&D, Dyson wants to automate flossing directly while you brush.
The Department of War just launched Starshield AI's Grok for Government on GenAI.mil, giving military personnel access to frontier reasoning models. Cleared for Impact Level 5 controlled unclassified data, the tool handles procurement research and supply chain workflows. GenAI.mil has already onboarded over 1.7 million users in nine months.
Google just introduced flexible, compute-based usage limits for Gemini Notebook, replacing the rigid daily reset with refreshes every five hours. The system factors in prompt complexity, chat length, and feature usage, and lets you queue heavy tasks like slide decks when running low. It rolls out to consumer accounts on web and mobile starting Sept 2.
ChatGPT and Codex just rolled out support for connecting multiple Gmail and Google Calendar accounts, tackling a long-requested headache. Users can now hook up second or third accounts directly via the settings gear next to connected Google services. Why it matters: You can finally query your work and personal schedules simultaneously without constant account switching.
Google just introduced Expert Intelligence, allowing users to import purchased ebooks from Google Play Books directly into Gemini Notebook. Launching with over 100,000 titles from publishers like Penguin Random House and O'Reilly, it lets you ask questions, generate audio overviews, and run quizzes grounded in the text. Crucially, shared collaborators must also own the book to access its insights.
Google and Khan Academy just rolled out new Gemini-powered features for AI tutor Khanmigo. The tutor can now generate interactive charts and geometry shapes on the fly as students work through STEM problems, adapting if they drag vertices or lines. A redesigned practice tool also lets teachers draft assignments with Gemini while keeping full editorial control. Districts using Khanmigo have access starting today.
Google just upgraded Search's AI Mode with three major travel features, allowing users to set flight price drop alerts across 300+ airlines, check reward points and miles redemption rates, and book hotels directly in chat via Google Pay. Why it matters: Google is steadily transforming its conversational search into an end-to-end travel agent that can handle everything from planning to checkout.
Plaud just launched the Plaud One, a $249.99 pair of AI earbuds designed to record, transcribe, and summarize real-world conversations. Its charging case features built-in 4G connectivity to sync notes and trigger workflows in apps like Slack and Notion without needing your phone nearby. Preorders are live now in select markets, with shipping kicking off in Q4 2026.
Adobe just rolled out a beta "AI Assisted Editor" for Photoshop, grouping its generative tools into a single streamlined toolbar. The biggest upgrade is a new markup tool that lets you draw arrows or brush rough shapes directly on an image to guide edits without typing a prompt, backed by Firefly Image 5 understanding full scene context.
Anthropic just opened a research preview of the Model Hardware Standard (MHS), an open protocol enabling AI agents to safely operate lab hardware like liquid handlers, microscopes, and robotic arms. Built with HHMI Janelia, MHS shrinks hardware integration from weeks to hours through standard drivers and MCP. Anthropic plans to open-source the spec after physical safety evaluations wrap up.
Google just upgraded Gemini Live with autonomous agent capabilities by integrating Spark, letting you trigger complex, multi-step tasks across Docs, Sheets, and Drive entirely by voice. The update also brings hands-free Gmail triage and a spoken Daily Brief. Why it matters: Voice assistants are finally evolving from passive Q&A toys into proactive agents that run errands in the background.
Particle just launched Radar, a podcast search engine indexing over 130,000 shows and adding 20,000 episodes daily. Designed to give AI agents access to audio they are usually blind to, it serves transcripts, speaker labels, and entity tracking via a search API and MCP. Hedge funds are already piling in as its highest-volume users.
Anthropic just added a built-in browser to Claude Cowork on desktop, letting the AI navigate sites, click, and fill out forms in an isolated side panel. It runs completely separate from your personal tabs, history, and passwords, requiring no extensions. Basically, you can now hand off tedious web chores to Claude without handing over your personal browser.
Anthropic just rolled out Claude in Chrome to all paid subscribers, letting the AI autonomously browse, click, and fill forms across tabs without pausing for manual approval on every step. A built-in classifier and probes scan for prompt injection attacks in real time, clocking a 0% attack success rate against Sonnet 5 and Opus 5.
xAI just expanded access to Grok Bot, unlocking its autonomous agent across all SuperGrok and Cursor paid tiers. Each Bot runs on its own always-on cloud computer with terminal and browser access to complete multi-app workflows. Best of all, Bot tasks come with dedicated usage that won't touch your existing limits.
ChatGPT just rolled out a new capability for ChatGPT Work that lets its built-in browser log into websites on web and mobile without ever seeing your username or password. Users can now delegate real-world errands like booking appointments, handling insurance reimbursements, or submitting invoices. It basically gives your AI agent hands to tackle authenticated web tasks without compromising security.
ChatGPT just upgraded scheduled tasks on web and mobile. Free users can now create up to three automated routines, while Plus and Pro subscribers get reactive triggers tied to changes in Slack, Gmail, and GitHub instead of running solely on fixed timers. Users can also share their favorite workflows for others to customize.
OpenAI just rolled out Premium seats for ChatGPT Business, aimed at power users hitting workflow bottlenecks. Priced at $125 per month (or $100 billed annually), it delivers 5x more usage than Standard seats, scraps the five-hour usage cap, and introduces predictable weekly resets. Why it matters: Teams can now mix and match standard and heavy-duty seats in one workspace without juggling multiple accounts.
