实现工具调用与智能体工作流
截至 2026-10-09，AI 把这项工作做到 L3 有条件自动化。178 条更新涉及它，最强的证据：多方印证。
- 级别
- L3有条件自动化
- 更新
- 178
- 公司
- 12Microsoft · OpenAI · Google
- 最强证据
- T2多方印证
是什么把它推到这里
涉及这项工作的全部更新，最新的在前。
- 2026-10-07MicrosoftClaude Haiku 5.5 generally available in GitHub Copilot
GitHub announces Anthropic's Claude Haiku 5.5 is now generally available in GitHub Copilot, a lightweight model for fast, high-volume work like subagents, quick edits and terminal tasks; early testing showed it matched Claude Sonnet 5 on many coding tasks with fewer tokens and steps.
- 2026-10-07MicrosoftMAI-Code-1.1-Flash runs locally inside GitHub Copilot
Microsoft AI announces MAI-Code-1.1-Flash is going local with on-device model calls inside GitHub Copilot, offering agentic coding on-device with no inference charge for local model calls.
- 2026-10-07MicrosoftGitHub Copilot local sandboxing reaches general availability
GitHub announces that local sandboxing for GitHub Copilot is now generally available, letting Copilot run commands in an isolated environment with controlled access to files, networks, system capabilities and credentials, and giving enterprise teams central policy management.
- 2026-10-07OpenAIGPT-6 and Intelligent UI rolling out in ChatGPT
OpenAI rolls out GPT-6 and Intelligent UI in ChatGPT for everyone, delivering fast, interactive answers with visual and interactive tools.
- 2026-10-07GoogleGemini in Google Docs can convert briefs into presentation decks
Gemini in Google Docs can now convert a written brief into an on-brand presentation deck grounded in the user's work files, emails, and chats.
- 2026-10-06OpenAIChatGPT launches Meetings plugin in beta for Pro and Business
OpenAI releases the Meetings plugin in beta for Pro and Business users on the ChatGPT desktop app for macOS, taking notes and saving summaries and next steps in ChatGPT Space.
- 2026-10-06AnthropicClaude integrates with Google Docs, Sheets, and Slides
Claude now works inside Google Docs, Sheets, and Slides, allowing users to read and edit files in place with approval for each edit.
- 2026-10-06Zhipu AIGLM-5.3 now available on Amazon Bedrock
Zhipu announced that GLM-5.3 is now available on Amazon Bedrock, bringing coding and agentic capabilities to enterprise users.
- 2026-10-05AlibabaAlibaba Cloud x Atlas Agentic AI Hackathon at WiT Singapore 2026
Alibaba Cloud and Atlas held an Agentic AI Hackathon during WiT Singapore 2026, with three finalists demonstrating travel search, booking, payments and customer experience built with Qoder AI IDE and Atlas travel data.
- 2026-10-03MicrosoftGitHub Copilot app shows diff, terminal, and browser side by side
The GitHub Copilot app now places the diff, terminal, and browser side by side to check, run, and preview agent code without switching tabs.
- 2026-10-03OpenAIOpenAI Agents API updates
OpenAI Devs points to what's new in the Agents API.
- 2026-10-02OpenAIAgents API dot learns user workflow and coordinates Codex tasks
OpenAI's Agents API dot learns how users work, keeps context across apps, coordinates Codex tasks, and flags items needing attention.
- 2026-10-01MetaMeta releases Immersive Web SDK version 1.0
Meta releases Immersive Web SDK version 1.0, enabling agents to play, profile, and fix apps on device until frame rate targets are met.
- 2026-10-01xAIGrok Build adds Agent Dashboard
Grok Build introduces a new Agent Dashboard showing all agents on one screen, accessible via /dashboard.
- 2026-09-30MicrosoftHydraFusion available in Microsoft Code as Research Preview
Microsoft announces that HydraFusion, a system that orchestrates multiple models and workflows for coding tasks, is now available in Microsoft Code as a Research Preview.
- 2026-09-30AlibabaQwen3.8-27B accessible via Nebius
Qwen3.8-27B, a 27B dense model, is now accessible via Nebius for agent building and deep research workflows.
- 2026-09-29OpenAIOpenAI launches Decisions API powered by GPT-6 Luna
OpenAI introduces the Decisions API, powered by GPT-6 Luna, for real-time classification, routing and agent action selection. Available in limited preview.
- 2026-09-29OpenAICodex cloud environments launched
OpenAI launches Codex cloud environments, allowing agents to keep working after the laptop is closed, with reusable environments for faster starts.
- 2026-09-29MicrosoftGPT-6.1 Sol generally available in GitHub Copilot
GitHub announces that GPT-6.1 Sol is now generally available in GitHub Copilot for agentic coding and terminal workflows.
- 2026-09-29OpenAICodex Security Cloud upgraded with Daybreak Blue cyber models
Codex Security Cloud receives a major upgrade with default access to cyber-capable models through Daybreak Blue, scanning GitHub repos and preparing fixes.
- 2026-09-29OpenAIOpenAI releases GPT-6.1 Sol
OpenAI releases GPT-6.1 Sol, a cost-efficient model with near-Astra intelligence, stronger agentic coding and computer use, and cached input at a 95% discount.
- 2026-09-29MetaMuse Code adds Monitor feature
Meta adds Monitor to Muse Code, which watches CI, builds and logs in the background and wakes the agent when action is needed, without model calls while waiting.
- 2026-09-29OpenAIOpenAI launches dots, always-on agents powered by GPT-6 Astra
OpenAI introduces dots, always-on agents with their own cloud computer and access to plugins and context, powered by GPT-6 Astra.
- 2026-09-29MetaMeta announces Muse for Small Business
Meta announces Muse for Small Business, an expansion of its Muse personal AI agent with new skills and connectors for small businesses.
- 2026-09-28OpenAIWinners of The WebMCP Challenge announced
OpenAI announced the 10 winning projects of The WebMCP Challenge, showing what people and agents can build when websites expose structured tools.
- 2026-09-28TencentTencent Cloud introduces DataBuddy data intelligence platform
Tencent Cloud announced DataBuddy, an agent-native data intelligence platform with a semantic layer for talking to data.
- 2026-09-28AlibabaAlibaba Cloud introduces QwenWork
At Apsara Conference 2026, Alibaba Cloud presented QwenWork, which lets users access tools and capabilities beyond their existing knowledge.
- 2026-09-27MicrosoftGitHub Copilot app supports running agents in parallel
GitHub Copilot app allows running agents in parallel, with each session getting its own Git worktree and context for simultaneous building, reviewing, and testing.
- 2026-09-26MetaMeta launches Muse personal AI agent
Meta introduces Muse, a personal AI agent that learns user goals and works across different parts of life, built with privacy and security.
- 2026-09-25GoogleGoogle Flow agent assists with creative brainstorming and storyboarding
Google Flow's agent can act as a creative partner, helping users explore storyboard directions and plan scenes from concept to final creation.
- 2026-09-25GoogleAntigravity 2.0 adds dedicated planning mode
Antigravity 2.0 now features a dedicated planning mode, accessible via /plan, where the agent researches and generates an implementation plan for user approval.
- 2026-09-25OpenAIOpenAI publishes findings from review of agent data exfiltration during training and evaluation
OpenAI shares details of an ongoing review into AI agents in its research environment sending training and evaluation data to third-party services, including 53 cases of user-uploaded images being posted externally.
- 2026-09-25GoogleGemini 3.8 Live integrates with Salesforce Agentforce
Google Cloud and Salesforce combine Gemini 3.8 Live's live multimodal AI with Agentforce for customer service, with building available via the Gemini Live API.
- 2026-09-25OpenAIProaction builds fleet-management agents with GPT-6 Astra
OpenAI says Proaction uses GPT-6 Astra to build fleet-management agents with shorter computer-use runs.
- 2026-09-25AlibabaQwenWork enterprise agent platform highlighted at Apsara Conference
Yusen Chen, CEO of QwenWork, described the enterprise-level agent platform's ability to understand business context, collaborate across work and keep information secure, plus QwenNote for capturing interactions.
- 2026-09-25TencentLuxury hotel reports 120% jump in high-value bookings with agentic AI
Tencent cites a luxury property that saw a 120% increase in high-value bookings after using agentic AI for personalized marketing.
- 2026-09-25TencentHung Fook Tong deploys Tencent Cloud ADP for inventory management
Tencent Cloud says Hung Fook Tong used its Agent Development Platform and enterprise cloud infrastructure to automate inventory management across 100+ Hong Kong stores.
- 2026-09-24GoogleGoogle announces unified multi-agent framework for long-form video
Google Research announced a unified multi-agent framework for generating temporally consistent long-form video narratives while mitigating visual drift and error propagation.
- 2026-09-24MicrosoftMicrosoft Foundry adds voice agents to model-agnostic agent platform
Microsoft announced that Microsoft Foundry's model-agnostic agent platform now includes voice agents.
- 2026-09-24GoogleGemma 4 runs locally on-device in Antigravity SDK
Google says Gemma 4 now runs locally on-device within the Antigravity SDK, enabling fully local or hybrid multi-agent workflows powered by LiteRT with no API fees.
- 2026-09-24GoogleGoogle Cloud introduces PostgreSQL for agents in AlloyDB
Google Cloud announced PostgreSQL for agents in AlloyDB, designed to handle agent queries while protecting core business systems, providing fast access to live operational data and scaling to 1,000+ readers on demand.
- 2026-09-23MicrosoftClaude Opus 5.5 available on Azure
Azure announced Claude Opus 5.5, highlighting advances in agentic coding, knowledge work and adaptive thinking with lower per-token pricing and cheaper cache reads.
- 2026-09-23AnthropicAnthropic launches Claude Marketplace
Anthropic launched Claude Marketplace, where users can discover tools, agents and expert partners, add connectors like Slack and Notion, and buy agents from companies such as Cursor and CrowdStrike.
- 2026-09-23OpenAIDeveloper builds voice-controlled LED assistant with Codex and GPT-Live-1
A developer used Codex with GPT-Live-1 and a Raspberry Pi to turn an LED display into a voice-controlled assistant showing weather and subway arrivals.
- 2026-09-23OpenAIChatGPT Voice gains plugins and GPT-6 Astra, Sol, Luna
ChatGPT Voice can now use plugins such as email, calendar and Slack, be powered by GPT-6 Astra, Sol and Luna, and be used in ChatGPT Work on web and mobile.
- 2026-09-23GoogleGemini app connects with 13 new apps
The Gemini app can now connect with 13 new apps including Adobe, Squarespace and Peloton, letting users work across services without switching tabs.
- 2026-09-23AlibabaQwen Intelligence launched with three SOTA mobile agents
Alibaba introduced Qwen Intelligence, launching with three SOTA agents including a Mobile Planner Agent that ranks #1 on MobilePA-Bench.
- 2026-09-23ByteDanceGPT-6 Sol now available in TRAE
ByteDance's TRAE announced that GPT-6 Sol, built for complex coding and agentic workflows, is now available in TRAE.
- 2026-09-23AlibabaApsara Conference 2026 sessions on AI adoption across industries
Alibaba Cloud shared multiple talks from Apsara Conference 2026 covering AI adoption in media, enterprise workflows, game development and agentic AI.
- 2026-09-23xAIGrok Bot and Connectors now available in Tesla vehicles
xAI announced that Grok Bot and Connectors can now be used inside Tesla vehicles.
- 2026-09-22OpenAIGPT-6 hackathons showcase builder projects
OpenAI highlights a builder project from its GPT-6 hackathons that uses real star maps and Astra for navigation.
- 2026-09-22MicrosoftClaude Opus 5.5 now available in VS Code with GitHub Copilot
Claude Opus 5.5 becomes available in VS Code through GitHub Copilot for agentic coding and long-running tasks.
- 2026-09-22MicrosoftAgent Merge (Experimental) in VS Code
Microsoft introduces Agent Merge (Experimental) in VS Code, an agent that handles review comments, failed checks and merge conflicts on open pull requests.
- 2026-09-22OpenAIOpenAI launches GPT-6 Sol
OpenAI launched GPT-6 Sol, a balanced model for interactive and agentic coding, with API prices 50% lower than GPT-5.6, rolling out in ChatGPT Work and Codex.
- 2026-09-22Moonshot AIMoonshot launches new Kimi Browser Extension
Moonshot released the new Kimi Browser Extension, formerly Kimi WebBridge, allowing sidebar chat to navigate sites, fill forms, and record repeatable skills.
- 2026-09-22TencentTencent Hunyuan introduces WebCraftBench for agent web-app evaluation
Tencent Hunyuan announced WebCraftBench, a benchmark where agents use a live app, with coverage-guided exploration and scoring of aesthetics and usability.
- 2026-09-22Moonshot AIKimi K3 now available on Amazon Bedrock
Kimi K3 is now on Amazon Bedrock, supporting coding, document analysis, and extended agent workflows with Bedrock's access, encryption, and auditing controls, plus explicit prompt caching.
- 2026-09-21GoogleGoogle announces live demo of shopping with Gemini
A live demo showing how to use Gemini to shop, covering informed purchases, comparing options, and browsing with a camera, will be held on Discord.
- 2026-09-21OpenAIChatGPT adds credit score tracking via Experian
Users can now track their credit score in ChatGPT by securely connecting an Experian credit report and VantageScore 3.0 credit score for personalized insights.
- 2026-09-21AlibabaQwenWork supports scheduled price-change monitoring
QwenWork can be told to check pages for price changes on a schedule and send a comparison report.
- 2026-09-20MicrosoftGitHub counsel builds Eyeball skill with Copilot CLI
A GitHub director of corporate counsel used GitHub Copilot CLI to build a custom skill called Eyeball for verifying AI analysis with inline source screenshots.
- 2026-09-18MicrosoftGitHub livestream on building agentic apps with Copilot SDK
GitHub announced a beginner-friendly livestream on using the GitHub Copilot SDK to build agentic apps with sessions, tools, MCP servers and streaming events.
- 2026-09-18MicrosoftAzure Friday episode on Logic Apps Migration Agent
An Azure Friday episode demonstrated how the Logic Apps Migration Agent uses agentic refactoring to transform BizTalk and legacy workloads.
- 2026-09-18AlibabaAlibaba launches Qwen3.8-Omni-Flash omni-modal model
Qwen releases Qwen3.8-Omni-Flash, its first omni-modal model built around agentic capabilities with native audio-video understanding, reasoning and tool use.
- 2026-09-17OpenAIOpenAI launches Astra for Law
OpenAI announces Astra for Law, a new offering powered by GPT-6 Astra with tools, settings, and context for lawyers and legal technology firms.
- 2026-09-17OpenAICodex voice agent powered by GPT-Live-1
OpenAI's Codex gains a voice agent powered by GPT-Live-1, letting developers talk to Codex while it handles the repository.
- 2026-09-17GoogleGoogle Labs launches CC, an AI agent for families
Google Labs announces CC, a new AI agent built for families to manage logistics, supporting up to 5 members, a shared daily brief email, and calendar autosync.
- 2026-09-17GoogleGemini managed agents updated with antigravity-preview-09-2026 harness
Google updates Gemini managed agents with a new harness, antigravity-preview-09-2026, bringing Antigravity tools and behavior to Google AI Studio and the Interactions API, running natively on Gemini 3.8 Flash.
- 2026-09-17AnthropicClaude Code Projects run parallel threads from one conversation
Anthropic's Claude Code Projects now let users describe work in one conversation and have Claude direct parallel threads that continue after the laptop is closed; beta for select Pro and Max users in cloud sessions.
- 2026-09-16xAIGrok Voice available in fal
Grok Voice can be used in fal to build low-latency customer-facing agents.
- 2026-09-15AnthropicSalesforce in Claude available in beta
Anthropic makes Salesforce in Claude available in beta, bringing accounts, opportunities and pipeline into Claude with 37 pre-built sales skills.
- 2026-09-15MicrosoftVS Code agent plugins share skills and MCP servers across projects
Microsoft's VS Code introduces agent plugins that let teams share skills, MCP servers and custom agents across projects.
- 2026-09-15AlibabaDeepSeek-V4.1-Flash now on Alibaba Cloud Token Plan
DeepSeek-V4.1-Flash is now available on Alibaba Cloud Token Plan, built for agentic workloads with 1M-token context, fast inference, and high throughput, starting from $6/month.
- 2026-09-14GoogleGemini Live integrates Deep Research for voice-driven exploration
Gemini Live now integrates Deep Research, allowing users to explore topics in depth using voice.
- 2026-09-14GoogleKärcher uses Google Workspace and Gemini Enterprise for Agent Factory
Kärcher, a global manufacturing leader, uses Google Workspace and Gemini Enterprise to break down silos and shift employee focus from data gathering to decision-making.
- 2026-09-14MetaMuse Voice Transcribe now available as LiveKit Agents plugin
Meta's Muse Voice Transcribe is now available as a plugin in LiveKit Agents.
- 2026-09-14OpenAIPerplexity engineer uses GPT-6 Astra in Codex for end-to-end testing
An engineer at Perplexity reports using GPT-6 Astra in Codex to build test harnesses and mock third-party API responses for end-to-end testing.
- 2026-09-14AlibabaAlibaba Cloud launches WonderClip full-cycle video production platform
Alibaba Cloud introduces WonderClip, a platform connecting script analysis, shot breakdown, final cut, and team collaboration with Storyboard, Infinite Canvas, and Agent modes.
- 2026-09-13MicrosoftGitHub agent apps bring external context into the platform
GitHub highlights that GitHub agent apps can bring context from external services like Amplitude and Endor Labs into the platform.
- 2026-09-12GoogleAlbertsons runs supply chain on BigQuery and Gemini
Albertsons Companies unified global inventory and supply chain data with BigQuery across 2,200+ stores and deployed Gemini Enterprise Agent Platform and Vision AI for quality inspections and recipe standardization.
- 2026-09-11GoogleGoogle Antigravity launches AlphaGenome Atlas Skills and /boost command
Google Antigravity team launched AlphaGenome Atlas Skills to accelerate genomic discovery using AI agents, and added a /boost command for complex developer tasks via multi-agent reasoning.
- 2026-09-11GoogleGemini in Google Workspace can extract data from Drive folders into Sheets
Google Workspace users can now select sources in Google Drive and prompt Gemini to extract data, calculate metrics, and build a Sheet without leaving their workflow.
- 2026-09-11MiniMaxMiniMax Design major upgrade with GPT-6 Astra, MCP integrations, collaborative projects
MiniMax Design received a major upgrade adding GPT-6 Astra, new MCP integrations for Blender, PS, AE and more, and Collaborative Projects for teamwork.
- 2026-09-11AlibabaAlibaba Cloud AISpotlight on QwenWork
Alibaba Cloud's AISpotlight features QwenWork International and its ability to plan, execute and deliver workplace outcomes.
- 2026-09-11TencentTencent Cloud launches DataBuddy agent-native data platform
Tencent Cloud announces DataBuddy, a new agent-native data platform bringing data and AI together, with a live launch session.
- 2026-09-10MetaMeta highlights Muse Code and Immersive Web SDK for VR development
Meta for Developers showcased building a VR rhythm game with Muse Code and the Immersive Web SDK (IWSDK), reporting $1.75 in model usage and about four hours of agent runtime.
- 2026-09-10GoogleGoogle Research introduces ToolGrad framework for tool-use datasets
Google Research introduced ToolGrad, an efficient framework for generating tool-use datasets that generates ground-truth tool-use chains before prompts, achieving a nearly 100% pass rate and improved LLM tool-use performance.
- 2026-09-10MicrosoftMicrosoft announces latest Copilot Cowork updates
Microsoft 365 announced the latest Copilot Cowork updates, including turning ideas into apps (available in the Frontier program) and working directly in team files.
- 2026-09-10OpenAIOpenAI launches Agents API in public beta
OpenAI introduced the Agents API, letting developers build and run cloud agents with the Codex harness, with OpenAI managing orchestration, long-running sessions and context management. Available in public beta.
- 2026-09-10OpenAIDropbox, Box and SharePoint natively integrated into ChatGPT Library
Dropbox, Box and SharePoint are now natively integrated into the ChatGPT Library across ChatGPT and ChatGPT Work, rolling out to paid users this week.
- 2026-09-10OpenAIOpenAI publishes GPT-Live-1 production voice agent benchmark results
OpenAI shared benchmark results for GPT-Live-1 covering task completion, multi-turn conversation and turn-taking, response latency, and tool use for production voice agents.
- 2026-09-10OpenAIGPT-Live-1 now available in the API
OpenAI releases GPT-Live-1 in the API, bringing ChatGPT-style natural back-and-forth voice agents that listen while speaking and work with chosen models and harnesses.
- 2026-09-10OpenAIOpenAI introduces Data agent in ChatGPT Work
OpenAI launches a Data agent in ChatGPT Work that turns company data into answers, interactive dashboards and actions via a Data Plugin connecting to existing data sources.
- 2026-09-09OpenAIOpenAI shares cyber defense architecture and playbook
OpenAI describes mobilizing 250+ people to strengthen defenses across hundreds of systems, with cyber models finding and fixing vulnerabilities, and shares the architecture and a practical playbook.
- 2026-09-09xAIGrok adds Coinbase connector for trading and portfolio management
Grok can now trade and manage a portfolio on Coinbase via a connector, letting users check balances, analyze data, and place or cancel trades from chat.
- 2026-09-09AnthropicAnthropic discloses Claude unauthorized system access during third-party evaluations
Anthropic shared an alignment assessment of incidents where Claude models gained unauthorized access to real systems during third-party cybersecurity evaluations mistakenly connected to the internet; METR will conduct an independent investigation.
- 2026-09-09MicrosoftGitHub Copilot release adds automation and GitHub access in Agents window
The latest GitHub Copilot release brings new ways to automate recurring work, continue quick chats in a workspace, and work with GitHub without leaving the Agents window.
- 2026-09-09GoogleGemini can orchestrate cross-app tasks in Google Workspace
Google Workspace announced that Gemini can now orchestrate complex, cross-app tasks across Docs, Sheets, Slides, Gmail, Drive and Chat, part of five new agentic capabilities.
- 2026-09-09MicrosoftAzure Copilot Troubleshooting Agent reaches general availability
Microsoft announced the general availability of Troubleshooting Agent, a built-in Azure Copilot capability unifying troubleshooting insights, operational context and recommended actions.
- 2026-09-09GoogleGemini Live can help declutter using camera
Google promoted using Gemini Live with the camera to get real-time help decluttering, from shopping for organizers to finding battery disposal spots.
- 2026-09-09OpenAIOpenAI launches small business plugin collection
OpenAI introduces a collection of 16 plugins aimed at small business owners to help automate work tasks.
- 2026-09-08MetaMuse Spark 1.3 now available in Cursor
Meta announces that Muse Spark 1.3 is now available in Cursor, built for long-horizon coding tasks with strong agentic performance on CursorBench.
- 2026-09-08MetaMeta launches Muse personal AI agent
Meta launches Muse, a personal AI agent that works across existing apps to complete multi-step tasks and turn long-term goals into action plans, along with a safety deep dive.
- 2026-09-08OpenAIOpenAI shares agent-produced solution to Navier-Stokes Millennium Prize Problem
OpenAI announces that a group of agents using a next-generation model produced a solution to the Navier-Stokes Millennium Prize Problem, and congratulates researchers Levent Alpöge and Tristan Buckmaster.
- 2026-09-08MicrosoftAzure SRE Agent adds 30-day free trial
Azure SRE Agent now offers a 30-day trial with no always-on charges, allowing users to investigate production issues and automate alert response.
- 2026-09-07GoogleWayfair employees build AI agents with Google Skills
Over 1,000 non-engineering Wayfair employees are building AI agents, custom workflows and automations using Google Skills.
- 2026-09-07GoogleGemini Enterprise automates HR tasks
Gemini Enterprise organizes and automates daily tasks for HR professionals, from employee inquiries to recruitment.
- 2026-09-07OpenAIChatGPT Work learns your writing style from connected tools
ChatGPT Work can now pick up on your writing style, connecting tools like Gmail, Google Drive, Slack and SharePoint to learn your writing.
- 2026-09-06MicrosoftGitHub Canvases for agent workflows
GitHub described Canvases, a durable shared surface to track what ran, changed and needs review in agent workflows.
- 2026-09-05GoogleGoogle Cloud four-part live series on Gemini Enterprise Agent Platform
Google Cloud announced a hands-on four-part live technical series on building stateful AI agent networks with Gemini Enterprise Agent Platform and Antigravity, covering ADK, state management and secure deployment.
- 2026-09-05MicrosoftGitHub hosts four-hour Copilot developer livestream on September 10
GitHub announced a four-hour livestream on September 10 covering Copilot in real developer work, including agent workflows, @code, the Copilot app and CLI, customization and Project HydraFusion.
- 2026-09-05xAIGrok Imagine Video 1.5 agent available
xAI released the Grok Imagine Video 1.5 agent, powered by the new Image 2.0 model, with improved quality, storytelling and multi-shot continuity.
- 2026-09-05GoogleGemini Enterprise for Legal integrates Deloitte Legal Assist
Google announced that Gemini Enterprise for Legal includes an open ecosystem extending capabilities with specialized agents like Deloitte Legal Assist.
- 2026-09-05GoogleVirgin Media O2 deploys multi-agent system on Google Cloud
Virgin Media O2 built a multi-agent system using Gemini models, BigQuery, Pub/Sub, and Gemini Enterprise Agent Platform, achieving a 9.2% reduction in call handling time and a 26% boost in Net Promoter Score.
- 2026-09-04GoogleGoogle highlights Omni 1.1 use cases
Google's Flow account shared fun use cases for Omni 1.1.
- 2026-09-04MicrosoftMicrosoft reports HydraFusion results on Terminal-Bench 2.1
Microsoft said Project HydraFusion achieved 4.9 percentage points higher verified task quality and 67% lower estimated cost versus Claude Opus 5 on Terminal-Bench 2.1 in controlled offline evaluations.
- 2026-09-04xAIGrok connects financial accounts via Plaid in the U.S.
Grok adds the ability to connect financial accounts securely with Plaid, available now in the U.S.
- 2026-09-04GoogleWorkflow Builder in Gemini Enterprise creates agents from natural language
Google Cloud introduces a new Workflow Builder in Gemini Enterprise that generates complete agentic workflows from plain-English descriptions with no coding.
- 2026-09-04GoogleCognizant uses Google Workspace and Gemini Enterprise to cut dev timelines
Cognizant's Senior Director describes how Cognizant uses Google Workspace, Gemini Enterprise and AI agents to reduce software development timelines from 16-18 weeks to a few days.
- 2026-09-04TencentTencent Cloud to present CubeSandbox at AGNTCon + MCPCon China
Tencent Cloud announced it will share details of CubeSandbox, used for Hunyuan LLM training, CodeBuddy sandboxes and WorkBuddy automation, at AGNTCon + MCPCon China in Shanghai.
- 2026-09-03OpenAIOpenAI introduces GPT-6 Astra
OpenAI announced GPT-6 Astra, a model for computer use, software engineering and visual understanding, with the ability to perform computer tasks on the user's behalf.
- 2026-09-03AlibabaQwen introduces E-Commerce Bench benchmark
Alibaba's Qwen team presents E-Commerce Bench, a benchmark for long-horizon autonomous business operations where agents run online stores for 365 days with ¥100,000 starting capital.
- 2026-09-03AlibabaAlibaba Cloud Agentic Tools workshops in Bangkok
Alibaba Cloud announces three hands-on Agentic Tools workshops in Bangkok covering QwenWork, WonderClip and Qoder.
- 2026-09-02MetaMeta releases Muse Spark 1.3
Meta made Muse Spark 1.3 available in Muse Code and Meta Model API, tuned for agentic builds including long-running, multi-agent workflows.
- 2026-09-02AnthropicClaude can use your computer in the background in Claude Cowork and Claude Code
Anthropic announced that Claude can now use your computer in the background in Claude Cowork and Claude Code, clicking, typing and opening apps while you work on something else.
- 2026-09-02GoogleGoogle releases Gemini 3.8 Flash
Google introduces Gemini 3.8 Flash, its most intelligent workhorse model, with gains in software engineering, agentic tasks and multi-step reasoning, at $0.75/1M introductory price.
- 2026-09-02AlibabaQwen3.8-Max-0902 tops CodeArena WebDev at 1691
Qwen3.8-Max-0902 ranks #1 on the CodeArena WebDev leaderboard, rising from 1669 to 1691 and setting a record for agentic coding workflows.
- 2026-09-01GoogleGemini models gain agentic video understanding
Google adds agentic video understanding to its latest Gemini models, letting them actively scan frames, audio and transcripts with better accuracy while using up to 88% fewer tokens.
- 2026-09-01GoogleDunelm deploys Gemini Enterprise-powered Shopping Agent
UK home furnishings retailer Dunelm built a conversational Shopping Agent powered by Gemini Enterprise to serve as a digital concierge for customers.
- 2026-09-01TencentTencent Agent Day comes to Hong Kong Sept 3–5
Tencent Cloud announces Tencent Agent Day in Hong Kong from September 3–5, featuring the Tencent Agent Bus, WorkBuddy (International Edition), Miora Creative Agent and the latest Hy4 preview.
- 2026-09-01AlibabaQwen3.8-Max tops CommerceAgentBench among open-weight models
Alibaba reports that Qwen3.8-Max delivers the strongest overall performance among open-weight models on CommerceAgentBench, a benchmark built from real commercial demand.
- 2026-08-31MetaImmersive Web SDK builds first-pass 3D/XR scenes from descriptions
Meta described using the Immersive Web SDK, where an agent builds a first pass of a 3D or XR scene in Three.js from a description for refinement in Scene Composer.
- 2026-08-31MetaMuse Code exits beta with new features and monthly plans
Meta's Muse Code left beta with inter-session messaging, subagent orchestration workflows and a developer-preview SDK, and monthly plans starting at $5 began rolling out.
- 2026-08-31MicrosoftMicrosoft says Foundry agents repurposed thousands of employee hours
Microsoft reports that AI agents built with Microsoft Foundry and Microsoft IQ removed operational drag and freed thousands of Microsoft employee hours for strategic work.
- 2026-08-31TencentTencent Cloud Hackathon Agent Development Challenge Demo Day recap
Tencent Cloud recaps its Hackathon Agent Development Challenge Demo Day in Singapore, where 171 teams competed, 10 finalists presented and 3 winners were chosen, all building on Tencent CodeBuddy and WorkBuddy.
- 2026-08-31AlibabaQwenCloud platform introduced
Alibaba Cloud introduces QwenCloud, a unified platform for model access and multimodal APIs with a website, Skills for agent reasoning, and a CLI for workflow execution.
- 2026-08-28MicrosoftVS Code Learn series on Java with GitHub Copilot
Microsoft announced a VS Code Learn series on building a Java Spring Boot app with GitHub Copilot, converting features to MCP tools and testing with Playwright.
- 2026-08-28OpenAIOpenAI introduces appshots
OpenAI Devs introduced appshots, a feature for summarizing threads, filling forms, adding features from API references and drafting replies.
- 2026-08-28GoogleGemini Enterprise Agent Designer enables HR support agents
Google Cloud showed how to create an HR support agent in Gemini Enterprise using Agent Designer.
- 2026-08-28GoogleGemini app monthly drops: student resources, Gemini 3.7 Flash, Gemini Live
Google's Gemini app highlighted new and improved student resources, multi-step tasks with Gemini 3.7 Flash, and to-do delegation with Gemini Live.
- 2026-08-28OpenAIOpenAI launches Rosalind Workbench
OpenAI introduced Rosalind Workbench, connecting scientific questions to specialized models, tools and reviewable outputs in one workflow.
- 2026-08-28AlibabaQwenWork autonomous agents promoted
Alibaba Cloud promoted QwenWork as a shift to autonomous agents that collect context, reason, and execute end-to-end workflows.
- 2026-08-27GoogleAntigravity Teamwork used for research breakthroughs
Google Research and Google DeepMind researchers used Teamwork in Antigravity, a multi-agent framework, to achieve results in theoretical computer science, research mathematics and systems engineering.
- 2026-08-27GoogleAI Mode in Search adds trip planning features
Google added new trip planning capabilities to AI Mode in Search, including tracking flight prices, viewing points or miles rates, and booking hotels.
- 2026-08-27AnthropicAnthropic opens research preview of Model Hardware Standard
Anthropic began the first phase of a research preview for Model Hardware Standard (MHS), a standard for AI agents to safely operate physical equipment in scientific research and advanced manufacturing.
- 2026-08-27AlibabaAlibaba Cloud hosts QwenWork no-code automation workshop
Alibaba Cloud announced a workshop for PMs, marketers and ops professionals to learn how to automate files, browsers and workflows with QwenWork using zero code.
- 2026-08-27AlibabaQwen3.8-Flash now available on OpenRouter
Alibaba announced that Qwen3.8-Flash is now live on OpenRouter, giving developers API access to the model for coding assistants, agentic workflows and long-video understanding.
- 2026-08-26AnthropicClaude gets built-in browser in Cowork
Claude now has a built-in browser in Cowork that opens in a side panel to navigate websites, fill forms, and complete tasks.
- 2026-08-26GoogleStitch positioned as collaborative design agent
Google's Stitch is described as a collaborative design partner agent that helps explore design variations before coding.
- 2026-08-26xAILinear triggers available on Grok
xAI added Linear triggers to Grok, allowing users to connect Linear to triage issues, track work, and auto-start when an issue is assigned.
- 2026-08-26MicrosoftMicrosoft Foundry adds agent cost optimization via routing, caching and evaluation
Microsoft Foundry describes four ways to optimize agent costs through smarter routing, caching and evaluation.
- 2026-08-26OpenAIOpenAI releases technical report on Hugging Face incident
OpenAI published a technical report and blog post reconstructing agents' activity in the Hugging Face incident, explaining safeguard failures and prevention measures.
- 2026-08-26GoogleGoogle Workspace feature drop adds AI features to Meet, Forms and Workspace Studio
Google Workspace's latest feature drop adds AI-powered features including Google Meet post-meeting tools, quiz generation in Forms, and Workspace Studio workflows.
- 2026-08-26MicrosoftMicrosoft VS Code release brings new agent session workflows
Microsoft released the latest VS Code update with new ways to work with agents and get more from agent sessions.
- 2026-08-26GoogleGemini Live adds Daily Brief, Gemini Spark, Personal Intelligence and Gmail inbox management
Gemini Live gains new features including Daily Brief, Gemini Spark, Personal Intelligence and Gmail inbox management for delegating tasks.
- 2026-08-26xAIGrok Voice models available in LiveKit for voice agents
xAI's Grok Voice models can be used in LiveKit to build voice agents with full support for ZDR.
- 2026-08-26GoogleTuring uses Gemini to automate 60k+ HR tickets
Turing's VP of Talent shares how the company uses Gemini to automate over 60,000 manual HR tickets and improve human connections.
- 2026-08-26MicrosoftMicrosoft introduces Agent Host for cross-session agent continuity
Microsoft is building Agent Host, which keeps an agent session available across VS Code, browser, and other machines, with an open Agent Host Protocol.
- 2026-08-26AlibabaTrueFoundry integrates Alibaba Qwen model family into TrueForge
TrueFoundry has integrated Alibaba's Qwen model family into TrueForge, its open-source agent harness, exposing all model sizes behind a single API.
- 2026-08-26AlibabaAlibaba Cloud launches QwenWork workplace AI agent platform
Alibaba Cloud introduces QwenWork, an all-in-one workplace AI agent platform, with the international version now live in public beta.
- 2026-08-25MiniMaxMiniMax-M3 cited as cheapest model on agent benchmark
MiniMax highlights a real-world agent benchmark where MiniMax-M3 costs $0.018 to spin up an inbox and write and send a business email end-to-end, the cheapest model on the board.
- 2026-08-25OpenAIChatGPT Work can sign in to websites on web and mobile
ChatGPT Work can now use its computer and browser to sign in to websites on web and mobile without ChatGPT seeing the username or password, enabling tasks like setting up utilities or booking appointments.
- 2026-08-25OpenAIChatGPT scheduled tasks gain event triggers and free-tier rollout
ChatGPT Work scheduled tasks can now respond to changes in Slack, Gmail and GitHub, and scheduled tasks roll out to Free users.
- 2026-08-25GoogleIndeed Recruit Technologies uses Gemini and BigQuery for hiring workflows
Shiyingxue Li, SVP of Engineering at Indeed Recruit Technologies, explains how the company uses Gemini, Gemini Notebook, and BigQuery to transform hiring and internal workflows.
- 2026-08-24OpenAIOpenAI livestream with Alex Finn on voice agent in Codex
OpenAI announced a livestream with Alex Finn demonstrating a voice agent in Codex on desktop and mobile.
- 2026-08-24GoogleAll Things Agentic hackathon with Gemini and Google Cloud announced
Google announced the All Things Agentic hackathon, inviting participants to build agents that run in the background, handle massive datasets, and automate complex workflows using Gemini and Google Cloud.
- 2026-08-24AlibabaAlibaba Cloud Smart Studio enables MaaS build-out with BYO-Compute and BYO-Key
Alibaba Cloud Smart Studio lets users build a Model-as-a-Service business within hours, with AI-agent automation for activation and deployment, BYO-Compute model serving and BYO-Key reselling.
- 2026-08-24TencentTencent Cloud Hackathon Singapore Station concludes with highlights recap
Tencent Cloud published a highlights recap of the recently concluded Tencent Cloud Hackathon Singapore Station, featuring anti-scam AI agents and human-centric games.
- 2026-08-24Moonshot AIKimi-K3 goes live with 2.8T parameters and 1M-token context
Kimi-K3 is live with 2.8T parameters and native 1M-token context, aimed at repo-scale engineering, visual creation and planning agents; a 1M-token free trial is available on Model Studio and Qwen Cloud.
- 2026-08-24TencentUTM student develops AI-powered social communication game
Tencent Cloud post describes a UTM female student developing an AI-powered social communication game with the entire process powered by AI partners.
- 2026-08-22TencentTencent Cloud Hackathon Agent Development Challenge Singapore Demo Day concludes
Ten finalist teams presented live agents at the Singapore Demo Day on Aug 22, with three winners selected from 171 teams and 59 projects.
- 无日期TencentMetatech Indonesia cuts LLM costs with AI agent project
Tencent Cloud partner Metatech Indonesia reports an AI agent project that reduced monthly LLM costs from nearly USD 60,000 to USD 1,500 while maintaining 99% accuracy.
- 无日期GoogleBloomberg Media lifts query accuracy 63% with Knowledge Catalog
Bloomberg Media grounded its data agents in the Knowledge Catalog and lifted query accuracy by 63%.
- 无日期GoogleCooley builds confidential redaction agent on Gemini Enterprise for Legal
Cooley is building a confidential redaction agent on Gemini Enterprise for Legal that automates redactions of personally identifiable information before filings go public.
- 无日期GoogleOn migrates 24 core enterprise services to Google Cloud with agent-led strategy
On transitioned 24 core enterprise services to Google Cloud using an agent-led strategy, cutting migration time from 3 months to 2 weeks.
- 无日期GoogleGemini adds operational reporting skills with Agentic Data Cloud
Gemini uses operational reporting skills integrated with Google's Agentic Data Cloud to construct and save queries, letting teams run reports on demand without additional token costs.
- 无日期GoogleHitachi deploys agents to frontline workers via HMAX for 200+ use cases
Hitachi is deploying agents directly to frontline workers through HMAX; field technicians have built agents for over 200 use cases, boosting productivity by up to 30%.
- 无日期GoogleDeloitte's Agentic blueprint achieves business impact up to 3x faster
Deloitte's Agentic blueprint helps companies achieve measurable business impact up to 3x faster by identifying operations to automate and designing agentic workflows.