Major AI platforms rolled out workplace integrations, new model tiers, and business-facing agents on November 10, while a critical inference-engine vulnerability exposed user data. Anthropic, Google, and OpenAI each expanded their enterprise toolkits, and Google Cloud added compliance features for regulated sectors.
Listen to today’s briefing (podcast)
Key takeaways
Anthropic adds Claude Meetings, Whiteboards, Sheets, and interactive connectors for Adobe, Atlassian, and Box to embed AI in live collaboration.
Google makes Gemini 4 Carbon generally available, topping independent benchmarks at half the per-task cost of rivals.
OpenAI launches ChatGPT Business Agents and GPT-Live models with full-duplex voice and delegation architecture for verified business interactions.
Google Cloud introduces Gemini for regulated industries with built-in governance, citations, and audit trails.
An RCE flaw in the open-source SGLang engine breached a NEAR AI inference server, exposing data from 18 agents and 100 users.
Anthropic launches Claude Meetings Whiteboards Sheets and interactive connectors
Anthropic announced new workplace features on Tuesday, November 10, 2026, including Claude Meetings, Claude Whiteboards, and Claude Sheets, according to TechCrunch. Claude Meetings brings Claude into video calls as a visible participant that can draft plans, map timelines on a whiteboard, answer questions in the meeting chat, and access connected work apps to retrieve information such as last year’s figures. After the call, meeting notes, action items, and other materials created by Claude are available in claude.ai. Anthropic says the feature turns discussions into actions in real time, whereas work usually starts after everyone hangs up.
Claude identifies itself as a meeting participant and posts a notice in the meeting chat. Claude Meetings will not join calls with external guests unless authorized by an admin, and will automatically leave if an external guest joins. The feature is launching in beta on Team and Enterprise plans, initially working with Google Meet, with support for additional platforms expected. Claude Whiteboards is a collaborative board where teams can talk through a plan and have Claude turn it into an organized visual workspace with timelines and deadlines. Teams can work directly with Claude on the same board in real time, eliminating the need to copy-paste results elsewhere, and when brainstorming is done, Claude can turn the board into a doc or a slide deck. Whiteboards also work inside Claude Meetings, so the board builds while people talk.
Claude Sheets lets teams create and collaboratively edit spreadsheets with Claude, including generating a working spreadsheet for a campaign tracker or hiring budget. Claude handles initial setup of trackers and budgets while allowing teams to continue refining and updating the same spreadsheet. Each spreadsheet can include multiple tabs, formulas, formatting, and dropdown menus. Claude makes edits directly in cells alongside the team and outlines each change so users can see what it did. Claude Whiteboards and Claude Sheets are available in beta on paid plans on web, desktop, and mobile. Claude Data (formerly known as Claude Dashboards) is now generally available, connecting Claude to systems like data warehouses or CRMs to answer questions with dashboards, written data stories, or narrated videos. Interactive connectors open the full version of apps like Adobe, Atlassian, and HubSpot inside Claude, allowing teams to open these apps directly alongside conversations and ask the assistant to work within them. Enterprise customers get security controls including audit logs, isolated virtual machines, and a compliance API, as reported by The Decoder.
Why it matters: These features integrate AI directly into meeting workflows and data tools, reducing context switching and enabling real-time automation of action items for IT teams and faster turnaround from discussion to deliverables for everyday users.
Google releases Gemini 4 Carbon generally available tops benchmarks at half the cost
Google has made Gemini 4 Carbon generally available to developers, enterprises, scientists, and consumers after a two-week early-access testing phase for selected security experts via the Fairwind Program, according to Google DeepMind’s blog and The Decoder. Google states that Carbon complements Gemini 4 Argon rather than replacing it; Argon remains the ideal choice for most software engineering and knowledge work, while Carbon is built for tasks that test the limits of a model’s intelligence like complex multi-step coding, debugging, and open-ended scientific research. Carbon uses more tokens than Argon to explore thoroughly, reason deeply, and run extended verification loops, so the total bill is likely higher despite the same token price.
Introductory pricing for Gemini 4 Carbon is $2 per million input tokens and $10 per million output tokens, scheduled to double to $4/$20 on December 1st, as noted by The Decoder, Simon Willison’s blog, SecurityWeek, and Google DeepMind’s blog. Paying developers and businesses can access Carbon via the Gemini API in Google AI Studio, Android Studio, Gemini for Business, and Google AI Ultra subscribers get access in Google Antigravity, the Gemini app, and AI Mode in Search. In independent benchmarks, Carbon ranks #1 on CursorBench 4.0, Perplexity’s WANDR, and Zapier’s AutomationBench at more than 2X lower cost per task compared to the next best external model. On the new Multimodal Knowledge Agent (MKA) benchmark developed with Artificial Analysis, Carbon ranks #2 while being more than 6X lower in average cost per task compared to the top-performing model.
In the Artificial Analysis test, Carbon cost $7.76 per solved task at the regular standard price of $4 per million input tokens and $20 per million output tokens, making it cheaper than the older budget model Flash 3.8 while delivering better performance. Google previewed the “/swarm” feature in Google Antigravity, which turns a single prompt into a coordinated team of agents tracked in real time, with Carbon handling deep reasoning roles like architecture design, coordination, and auditing while faster models run in parallel as adversarial verifiers. Demonstrations include an open-world 3D game exploring Google’s Mountain View campus as a pelican on a GBike, an interactive 3D reconstruction of the Eiffel Tower’s 1887-1889 construction from historical photos, and a playable 18th-century fortepiano with seat-specific spatial acoustics. In cybersecurity, Carbon found real zero-day vulnerabilities while scanning two large SaaS providers in under two hours as part of Wiz’s Scan for Good program, and Google’s Cloud Vulnerability Research team used it to write working proof-of-concept exploits for two known bugs in a popular open-source project with severity comparable to the PHP2Shell vulnerability. In Google’s Project Fortify, Carbon boosted the detection rate on open-source codebases by more than 25 percent over Argon.
Why it matters: Gemini 4 Carbon’s general availability brings a high-performance model with leading benchmark scores at half the per-task cost of competitors, enabling cost-effective deployment of advanced AI agents for coding, research, and cybersecurity.
OpenAI launches ChatGPT Business Agents and GPT-Live for verified business interactions
OpenAI is launching a beta feature called “Businesses in ChatGPT” that could compete with Google’s “MyBusiness,” according to The Decoder. Companies can register and manage their business through a Business Hub, update information, add product feeds with prices and availability, and track how often they’re mentioned in conversations or how many clicks their website gets. Eligible US e-commerce and online service providers can create “Business Agents,” AI agents that answer customer questions in ChatGPT, help with product comparisons, and can serve as targets for ChatGPT ads. Shopify merchants can connect their store directly. OpenAI says user conversations aren’t shared with businesses. ChatGPT currently reaches 1.2 billion users per week, according to OpenAI, and the company plans to expand the offering to more markets and industries over time.
GPT-Live is a new series of models with two key architectural changes: full duplex (can listen and speak at the same time) and delegation (splits conversation handling from complex work delegation to backend models), as shown in OpenAI’s YouTube presentation. GPT-Live can better understand acknowledgments versus interruptions, making conversations more natural. Delegation allows swapping newer backend models as they become available and adding voice to existing text applications. GPT-Live handles multiple people and busy environments by leveraging conversation context to decide when to interject. Developers can influence behavior via prompting, e.g., “ignore any side conversations and only respond when you’re spoken to.” Yelp uses GPT-Live for restaurant reservations and food orders; Fin for customer support; Speak for live language lessons; Cognition for AI software engineer Devin. New voices released include two French voices: Brise and Sillage. Custom voice generation via prompt-to-voice (natural language description) or from recording (with speaker consent) is also available.
ChatGPT is getting verified “Business Agents” that represent brands for comparison shopping, available first to eligible e-commerce and online service businesses in the US as part of beta business-focused features, according to The Verge. Users can start talking with a Business Agent if a business that created one appears in a ChatGPT response. Chats with Business Agents are clearly labeled and separate from main ChatGPT conversation. Companies can claim organization and get verified through Business Hub; OpenAI not sharing specifics about verification methods (different methods for different businesses) per Benji Shomair, VP of monetization. Business Hub shows insights: mentions in ChatGPT responses and top “discovery topics.” On Tuesday (Nov 10, 2026), OpenAI rolled out Business Hub for companies to create business profiles to verify and update information about products/services; feature in beta, as reported by TechCrunch. OpenAI previously tried ChatGPT as e-commerce hub with commission fee system and checkout cart, retired in March. OpenAI blog states: “Manage your business information, get insights about how your business shows up, and turn interest into action in ChatGPT.”
Why it matters: The new Business Agents and GPT-Live delegation architecture provide programmable, low-latency voice interfaces that integrate with existing backend systems, enabling scalable AI-driven customer interactions for IT professionals and trustworthy conversational access to products and services for everyday users.
Google Cloud launches Gemini for regulated industries with cited and auditable AI
Google Cloud’s Gemini app comes with governance, identity management and auditability built in for regulated industries like banks and capital markets, according to the YouTube video ‘Gemini for regulated industries: Cited & auditable AI’. The agent provides auditability and traceability throughout the workflow. Users get citations and methodology in everyday work, enabling confidence in Gemini’s analysis.
Why it matters: This enables AI deployment in highly regulated environments with built-in compliance, governance, and audit trails for IT professionals, while giving everyday people more transparent, verifiable decision-making from financial institutions using AI.
SGLang RCE exploit breaches NEAR AI inference server steals data from 18 agents and 100 users
An attacker exploited a remote code execution flaw in SGLang, an open-source inference engine, and ran code on one of NEAR AI’s inference servers last Sunday, according to The Hacker News. The attacker could also have seen requests being processed on that server that day, including requests sent in confidential mode. Confidential mode processes requests inside a trusted execution environment (TEE), but the RCE bypassed that protection.
Why it matters: This incident highlights the critical risk of remote code execution vulnerabilities in open-source AI inference engines and the potential bypass of trusted execution environments, exposing confidential data processed by AI services.
MeetGeek — MeetGeek provides automated meeting recording, transcription, and summarization that complements Claude’s new meeting integration capabilities. Try MeetGeek to automatically capture and summarize your meetings with AI-powered insights. Try MeetGeek
Runpod — Developers exploring Gemini 4 Carbon’s capabilities need scalable GPU infrastructure for fine-tuning and deployment. Runpod provides on-demand and serverless GPUs to train, fine-tune, and deploy AI models without managing infrastructure. Try Runpod
Cybersecurity Book on Amazon — A cybersecurity book helps developers understand and prevent inference-server exploits like the SGLang RCE. See current options and reviews on Amazon. View on Amazon
Affiliate disclosure: some links in this description are affiliate links. If you sign up through them, SysArticles may earn a commission at no extra cost to you. We only mention tools that are relevant to the story. As an Amazon Associate I earn from qualifying purchases.
Today’s releases show AI platforms deepening enterprise integration while a serious inference-engine breach underscores the operational risks of open-source components. Expect compliance tooling and agent interoperability to remain focal points through Q4.
This briefing is produced with AI assistance from the linked primary sources. Narration uses a synthetic AI news-anchor voice, built with Higgs Materials licensed from Boson AI USA, Inc.
Ali YAZICI is a Senior IT Infrastructure Manager with 15+ years of enterprise experience. While a recognized expert in datacenter architecture, multi-cloud environments, storage, and advanced data protection and Commvault automation , his current focus is on next-generation datacenter technologies, including NVIDIA GPU architecture, high-performance server virtualization, and implementing AI-driven tools. He shares his practical, hands-on experience and combination of his personal field notes and “Expert-Driven AI.” he use AI tools as an assistant to structure drafts, which he then heavily edit, fact-check, and infuse with my own practical experience, original screenshots , and “in-the-trenches” insights that only a human expert can provide.
