All news

TThurrott09:26Product release

AI score53

Opera launches WebAI local AI preview in Opera One, GX, and Neon

Opera announced that its WebAI local AI capability has entered early preview in Opera One, Opera GX, and Opera Neon, supporting the Prompt, Decision, and Semantic Embeddings APIs, allowing developers to call various models directly within the browser, including open-weight models, without installing or configuring AI applications. Opera also provided six experimental extension examples, such as Project Lens, which colors article paragraphs by project relevance, and Got It?, which generates questions about the article being read. The feature is opt-in and can be disabled in settings; currently Opera One is available on Windows x64 and Mac Apple Silicon, with GX and Neon to follow later.

Same story · 5 reports

#Product updates#English text is being prepared.#English text is being prepared.

UUnite.AI09:26Product release

AI score60

Oracle and OpenAI Expand Partnership, Launch Four Enterprise AI Products

At Oracle AI World in Las Vegas, Oracle announced an expanded partnership with OpenAI, launching four products—Oracle AI Workspace, Oracle AI Security Mesh, Oracle App Composer, and Oracle AI SMB Bundle—that combine OpenAI models, ChatGPT, and Codex with Oracle applications, data, and cloud infrastructure. Among them, AI Workspace connects to Fusion Applications, NetSuite, Oracle AI Database, and OCI through a ChatGPT plug-in and is coming soon; App Composer lets employees build enterprise applications with Codex, with Oracle managing deployment on OCI, and is currently in Limited Availability. AI SMB Bundle is aimed at eligible U.S. private small businesses, providing $12,000 in annual Oracle cloud credits for up to three years.

Same story · 6 reports

#Industry news#OpenAI#Agents

UUnite.AI09:26Product release

AI score59

Mozilla adds AI form filling and page monitoring to Firefox Smart Window

On October 27, 2026, Mozilla announced that the beta version of Firefox's Smart Window adds two capabilities: extracting information from user-selected tabs to help fill out online forms, and monitoring web page changes based on user descriptions. Form filling uses Gemini 3.1 Flash Lite to identify fields and extract information, while page monitoring uses the Qwen model, even if another model is selected in the Smart Window settings; saved autofill data remains local and is not sent to the model. Both features are currently available to English-language desktop beta users in the United States and Canada, and page monitoring tracks up to 5 pages per task and retains up to 15 active tasks.

Same story · 3 reports

#Product updates#Agents#English text is being prepared.

UUnite.AI09:26Product release

AI score59

Oracle Releases AI Security Mesh Powered by OpenAI Cybersecurity Models

At Oracle AI World, Oracle released Oracle AI Security Mesh, a managed platform powered by OpenAI's frontier cybersecurity models that helps enterprises discover and remediate security vulnerabilities faster, covering applications and infrastructure on-premises, on OCI, and on other clouds. The platform first uses a discovery framework to map customer systems, then dedicated agents investigate weaknesses, verify reproducibility, and help prepare code fixes; all fixes must be approved by the customer before being applied, and each agent runs with an independent identity and just-in-time permissions. Oracle also mentioned that OCI Identity Assurance can use biometrics to confirm the identity of approvers for sensitive operations, while OCI Zero Trust Packet Routing provides underlying network access control for AI workloads.

Same story · 6 reports

#Product updates#Safety / alignment#Agents

UUnite.AI09:26Product release

AI score52

Brilliant Labs launches $399 Halo AI glasses, focusing on persistent memory and adaptive gaming

Brilliant Labs has launched the $399 Halo smart glasses, built around the Noa assistant and an adaptive game called Narrative, divided into three experience categories: Ask, Make, and Play. The hardware includes a full-color microOLED display, dual microphones, a bone conduction speaker, and Bluetooth 5.3, weighs slightly over 40 grams, and the company claims up to 14 hours of battery life, powered by the Alif Balletto B1 processor and Ethos-U55 NPU. Noa and miniapps include daily usage limits; Noa Plus costs $9.99 per month and Noa Pro costs $39.99 per month. The company also announced an investment from Goodwater Capital, without disclosing the amount.

Same story · 2 reports

#Product updates#English text is being prepared.#Speech

MMacRumors Apple Intelligence09:26Product release

AI score62

ChatGPT Adds iCloud Mail and Calendar Plugins

OpenAI says it has added iCloud Mail and iCloud Calendar plugins to ChatGPT. The calendar integration can read, schedule, and update events in the iPhone calendar and in third-party apps that use iCloud Calendar; the mail integration can search and read iCloud Mail, manage emails and drafts, send emails, and look up iCloud contacts. ChatGPT can only access iCloud mailboxes at @icloud.com, @me.com, and @mac.com, and cannot see other accounts such as Gmail or Outlook in the iOS Mail app, and both integrations require creating an app-specific password. The plugins are available across all plans on web, desktop, and iOS, and can also be used with Dots, the proactive agent OpenAI launched last month.

Same story · 5 reports

#Product updates#OpenAI#English text is being prepared.

CCNET09:26Product release

AI score51

Apple Creator Studio Gets AI Update, Final Cut Pro Built-in MotionVFX

Apple announced that its Apple Creator Studio series of creative software will receive multiple AI updates. Final Cut Pro will have MotionVFX built in, bringing more visual and graphic effects tools to Mac editors; Logic Pro is launching a new AI assistant beta for U.S. music editors, powered by models built by Apple and OpenAI, and features a browser redesigned in the Liquid Glass style; Pixelmator Pro will gain a generative fill-like tool that can use AI to resize and fill in blank areas of new images.

Same story · 9 reports

#Product updates#Video#Speech

RReplit Blog09:25First-partyProduct release

AI score70

Replit Cloud Agents are now available to Pro and Enterprise users…

Replit announced that Cloud Agents are now available to all Pro and Enterprise customers. Non-engineers can describe changes in natural language, and the agent locates the code, runs the application in a cloud sandbox to verify it, and submits a PR or MR with screenshots and video to GitHub or GitLab. Teams can configure a shared environment once, and review feedback can be turned into skill updates that require engineer approval; the feature is in early access. The Slack integration is now available to all customers, and Microsoft Teams support is coming soon.

#Product updates#Agents#English text is being prepared.

TThe Verge AI09:24Product release

AI score50

Firefox Smart Window adds AI form filling and web page monitoring features

Mozilla has added a form filling feature to Smart Window, Firefox's AI browsing experience. It can extract information from open tabs to generate filling suggestions, and the data will be shared with Google's Gemini 3.1 Flash Lite model. Users can also type /watch in the Smart Window chat to have the browser monitor specified web pages daily or weekly for changes such as promotions and job postings. It only checks while the browser is open, pauses after 60 days without changes, and stops after 90 days. Both features have launched in the Smart Window beta on desktop in the United States and Canada, and will roll out in other countries in the coming months.

Same story · 3 reports

#Product updates#Agents#English text is being prepared.

IEnglish text is being prepared.09:24Product release

AI score60

Opera Launches WebAI Web Standard, Enabling AI Models to Run Locally in the Browser

Opera has launched a new web standard called WebAI, allowing users to run AI models directly in the browser without relying on cloud services. The standard is not tied to any specific model; taking Google's open-weight model Gemma as an example, on Mac it also supports calling Apple's on-device foundation models. It is currently available as an early preview in Opera One and requires users to enable it manually, with Opera GX and Opera Neon to follow. Opera says its advantages lie in privacy protection and no subscription, but actual performance depends on the device's hardware configuration.

Same story · 5 reports

#Product updates#English text is being prepared.#Open-source ecosystem

IEnglish text is being prepared.09:24Product release

AI score67

OpenAI opens iCloud Mail and iCloud Calendar plugins to all ChatGPT plan users

On October 27, OpenAI published a blog post opening two plugins, iCloud Mail and iCloud Calendar, to all ChatGPT plan users, covering consumer plans such as Free, Go, Plus, and Pro, and supporting the web, desktop, and iOS versions. After user authorization, ChatGPT can read and check the iCloud inbox, create and manage emails, and create or update events in Apple Calendar based on meeting, flight, and schedule information in emails. AppleInsider noted that the update has sparked discussion on privacy and competition, and that sensitive account, travel, financial, and work information in emails has made permission boundaries and data processing methods a focus of user concern.

Same story · 5 reports

#Product updates#OpenAI#English text is being prepared.

CClaude Blog09:23First-partyProduct release

AI score61

Anthropic Launches Claude for Financial Advisors Connectors and Skills Suite

Anthropic has released Claude for Financial Advisors, connecting data from custodians, asset management and wealthtech platforms into Claude, and providing skills tailored to financial advisors' daily work. The suite includes connectors such as Addepar, BlackRock, Charles Schwab, Envestnet, FactSet, Orion and Wealthbox, as well as skills for meeting preparation, portfolio rebalancing review, post-meeting notes and follow-up, and compliance and AI policy checks; regulated steps such as investment advice and client communication still require human review. The advisor plugin is available starting today, and the official recommendation is for registered investment advisors to use the Enterprise plan with audit logs.

#Product updates#Agents#English text is being prepared.

SShanghai Observer09:22Product release

AI score63

AgiBot Releases Heavy-Load Embodied Robot Genie G2 Max, Already Handling Joint Transport on the Qiyuan T1 Production Line

AgiBot has released the safety-certified heavy-load embodied robot Genie G2 Max, which has already taken on the task of transporting joint components on the Qiyuan T1 production line, where one robot rolls off the line every 2.5 minutes. The G2 Max has a single-arm payload of 20 kilograms and a dual-arm peak payload of 50 kilograms, with repeat positioning accuracy within 0.5 millimeters; it supports dual-battery hot swapping and autonomous recharging, enabling 7×24-hour operation. AgiBot also released the embodied intelligence agent framework Genie Master at the same time. The Genie G2 equipped with this framework has already started work at a convenience store in Shanghai, where it can autonomously identify the cause of failure and adjust its strategy in an unmanned environment.

Same story · 7 reports

#Industry news#Embodied AI#Agents

TTechSpot09:22Product release

AI score65

Open source project gpu-unlock unlocks the AI compute limits of RTX 20/30/40 series graphics cards

GitHub user Zagitta released gpu-unlock, which rewrites the GPU's speed select register before the Nvidia driver loads, removing the half-speed limit on FP32 accumulation tensor operations on GeForce graphics cards, without the need to flash the VBIOS or modify the driver. On a 450W RTX 3090 Ti, BF16 throughput increased from 87.5 to 157.4 TFLOPS, TF32 nearly doubled, and FP16 improved by 35%; the Ada-compatible gpu-unlock-ada raised the RTX 4090's FP16 instruction rate from 165.6 to 331.6 TFLOPS, but independent testers measured real-world matrix workload gains of 12% to 34%. turing-unlock targets the RTX 2080 Ti, making FP16 matrix operations 36% to 64% faster. It currently mainly supports Linux; gaming and chatbot token speeds are basically unaffected, prompt processing is 14% to 36% faster, and image generation and fine-tuning benefit the most.

Same story · 2 reports

#English text is being prepared.#English text is being prepared.#English text is being prepared.

CcnBeta09:22Product release

AI score54

Firefox tests AI Smart Window feature that can automatically fill out web forms

Mozilla demonstrated Firefox's Smart Window feature, which runs as a sidebar and lets users use natural language to ask it to fill out web forms such as job applications, service bookings, or online registrations. The feature analyzes form fields on the page, extracts names, contact information, and work experience from materials such as resumes, and fills them into the corresponding fields. Before submission, it shows the user the content to be submitted for review and modification. Mozilla says its positioning is as an intelligent assistant that can understand and operate web content, rather than a mere chatbot, and emphasizes that users can control the scope of information the AI can access. The feature is currently still in testing, and no timeline has been announced for making it available to all Firefox users.

#Product updates#Agents

CcnBeta09:22Product release

AI score67

Apple adds AI assistant to Pages, Keynote, and Numbers

Apple has added an AI feature called "Smart Assistant" to Pages, Keynote, and Numbers, powered by models from Apple and OpenAI, capable of drafting articles, generating slides, writing speaker notes, generating formulas, and recommending charts. The feature launches in beta, requires iOS 27, iPadOS 27, or macOS 27 or later with Apple Intelligence support, and availability is subject to language and subscription conditions. Apple has also begun enabling MCP support for these three apps in macOS 27, allowing eligible third-party AI agents to connect and operate documents, presentations, and spreadsheets.

Same story · 9 reports

#Product updates#Agents#English text is being prepared.

TTogether AI Blog09:22First-partyProduct release

AI score52

Together AI Details GPU Cluster Auto-Repair: From GPU Failure to Healthy Node

Together AI introduces its GPU cluster auto-repair system, which now officially offers passive health checks, repair recommendations, auto-repair policies (including per-fault and per-node authorization), wait-for-idle controls, escalation ladders, safety guardrails, and API and console repairs; custom detection onboarding based on node labels remains in preview. The system is divided into four stages: detection, decision, repair, and re-verification. The first three stages run within the user's cluster, while re-verification is performed on the Together side on the freed-up physical machines. The platform maps each signal to the lightest repair action, and users can decide whether to execute automatically and how long to wait; after the same action is repeated twice for the same fault, it automatically escalates one level.

#Product updates#English text is being prepared.

OOpenRouter：Announcements08:00First-partyProduct release

AI score62

OpenRouter Rebuilds Provider Routing Algorithm, Reducing TTFT by 16.9%

OpenRouter announced that it has rebuilt its provider routing algorithm, shifting from allocating traffic mainly based on input token price to a real-time routing cost that comprehensively factors in actual price, time to first token (TTFT), throughput, error rate, and output quality. The company says the new algorithm reduces TTFT by 16.9%, increases throughput by 16.3%, and cuts 429 errors by 46%, while cost per request rises by 6.0%. Users do not need to change any code; explicitly setting provider.sort or provider.order can still bypass the new logic.

#Product updates#English text is being prepared.#Reasoning

CClaude Blog08:00First-partyProduct release

AI score55

Anthropic Launches Legal Work Artifacts Library for Claude for Legal

Anthropic has launched directly usable legal artifacts for Claude for Legal; teams can adapt them with their own data and connectors and share them across departments. The legal gallery covers tasks such as contract review, regulatory change tracking, and business legal requests; each artifact comes with examples, and all outputs include source links. Since launching in May, more than 60 legal connectors have been added, including Bloomberg Law, Filevine, Ivo, Legora, and Patlytics.

Same story · 3 reports

#Product updates#Anthropic#Agents

GGoogle Blog08:00First-partyProduct release

AI score57

Google Launches Connected Apps Platform, Letting Developers Integrate with Search and Gemini Through a Unified Interface

Google has released the preview version of the Connected Apps Platform, where developers can build and manage Connected Apps for Search and Gemini applications on the same platform. Developers complete integration through a cloud-hosted MCP server or Android MCP combined with AppFunctions APIs, then use the Google Connected Console to register, test, and submit, and choose to launch on Search, Gemini, or both; after launch, they can view traffic, latency, and error rates. Google says Search and Gemini already have more than 50 Connected Apps, with integrations from Adobe, Experian, Localiza, and Peloton already live on Gemini and set to arrive in Search in the coming weeks, while Calm's integration will also launch on both.

Same story · 2 reports

#Product updates#English text is being prepared.#Google