Korea Aerospace Industries kicked off a project to develop an AI agent combining AI with military logistics big data, at a Daejeon meeting with 100 participants including the Ministry of National Defense.
Crypto trading platform PrimeXBT lets users connect ChatGPT and Claude to their trading accounts through an MCP integration, enabling AI-assisted trading actions and queries.
HtAG is exposing Australian commercial property sales data to AI assistants through an MCP server, letting agents query the dataset directly.
Otto launched a voice AI CRM that lets field sales representatives update sales records with a phone call.
Chrome update addresses 213 vulnerabilities including 7 critical; 14 reports aided by AI.
Pika's H3 Flex video generation model is pitched as giving users finer control over AI-generated video output. Specific controls, pricing and availability are not detailed in the source.
NetEase Yidun's Agent Guard checks agent intent, permissions and tool calls rather than only outputs, alongside Skill and MCP supply-chain scans, the company said at Wuzhen.
Sunra is an AI media workspace for generating and working with images and video. The source does not detail pricing, model stack or availability.
Google's Dreambeans app creates personalized stories for users 18 and older.
CrowdStrike is offering enterprises a way to instantly halt misbehaving AI agents, part of its push to secure agentic AI deployments with kill-switch style controls.
Why it matters: Adds a major Chinese lab to the open-weight race with DeepSeek and Qwen.
Tencent released Hy4, an open-weight AI model aimed at breaking into the top tier of open models alongside DeepSeek, Z.ai and Alibaba's Qwen. One report says Astra fine-tuned Qwen weights for longer reasoning.
Hydrogen-1.0 is a scientific problem-solving and research assistance model built on GLM-5.3 for explaining derivations and organizing research answers.
Alibaba Cloud launches Qovest, an AI investment research workbench; individual invite-only testing opens at 99 yuan per month.
Why it matters: Extends Mistral beyond LLMs into efficient time-series forecasting without GPU dependence.
Mistral launches the four-model Timestral family, which forecasts related data streams and planned inputs and can run without a GPU; results submitted to GIFT-Eval and TIME.
BluePixel-Turbo is a model for native 4K instruction-based image editing that works in four steps, shared on Reddit's StableDiffusion community.
South Korea's KCC is showcasing automotive refinish coatings and AI-powered color-matching technology at SEMA Show 2026 in Las Vegas to grow its North American aftermarket presence.
Eney is an AI assistant from MacPaw, the maker of CleanMyMac and Setapp. It runs on Mac and iPhone and plugs into Apple services like Mail, Calendar and Reminders.
Why it matters: A high-profile design firm restructuring around AI, signaling pressure across creative software.
Canva will cut or relocate more than 1,200 jobs, about 22% of its workforce, in a sweeping overhaul it says AI demands, with reports citing rollout problems.
Idol Turbo, a new image generation model checkpoint, has been published on Hugging Face by akhaliq. The source gives no parameter count, license or benchmark details.
A llama.cpp pull request by max-krasnyansky adds graph optimizations for Gemma 4 E2B and E4B, reporting large prompt-processing speedups for the small open models.
Why it matters: Shows the tension between falling inference prices and massive compute costs at a leading lab.
Anthropic swung to a Q3 operating loss after cutting AI model prices, even as Claude's revenue far outpaced Chinese rivals Zhipu and MiniMax; reports cite a $66.3B net loss and an $895B compute burden ahead of a possible IPO.
Uncracked City is a detective game where players interrogate suspects powered by large language models that can lie. It was posted on Show HN.
UK startup Mediacendo launched an AI operating system for paid media with initial funding. No amount was given.
Why it matters: Extends coding agents into full game-engine project editing with long unattended runs.
Kimi's game development solution pairs K3.1 models with Kimi Code and Kimi Work agents that edit Unity, Unreal Engine and Blender projects, including a 20-hour unattended run.
Soniox releases STT v6, a speech-to-text update for voice agents that adds agent text input and echo filtering, aimed at making real-time voice agent conversations cleaner.
Why it matters: Gives a cryptographic way to prove model identity without revealing weights, a building block for agent trust.
FlashProve is an open-source 'provable inference' tool that works like a cryptographic signature for AI outputs, letting users verify a response came from the claimed model without exposing its weights.
Why it matters: Opens Apple's assistant to third-party app actions, a key step for agentic use on iPhones.
Apple's iOS 27.3 beta gives Siri AI access to third-party apps through a developer preview for custom app integrations, and iPadOS 27.3 extends Siri mode to the Camera app.
Walmart's AI shopping agent Sparky can now use user-uploaded images to answer questions or build shopping lists, among several new capabilities added ahead of the holiday season.
Keytomic released an MCP server, an AI visibility tracker and a Reddit AI agent aimed at automating SEO workflows for marketers.
Agoda has deployed an assistant built with OpenAI to help users with travel search and booking tasks. The source gives no further product or availability details.
Itch.io added a filter letting users find or hide games that disclose generative AI use, a step Steam has not yet taken.
Anthropic now lets Claude automatically connect 14 workplace plugins when Team and Enterprise users sign in, vouching for them where their identity provider can't support enterprise-managed authorization.
Why it matters: India's state-backed push into sovereign open foundation models.
IndiaAI Mission released indigenous foundation models EKA-1, Dhīr 2B and Dhīr 5B, built by Soket Labs and ZenteiQ, targeting offline mobile access and secure use in critical sectors. Sarvam also launched an open-weight decision model.
Cohere says its leading open AI models are now available in Model Vault; the source gives no further details on which models or availability terms.
Why it matters: Scale of off-balance-sheet AI commitments is a key risk indicator for the whole buildout.
Nikkei Asia reports US tech giants' hidden AI debt rose 90% in six months to $3.2tn, raising questions about how labs like OpenAI and Anthropic will pay for compute given weak productivity gains.
Cyclon is a new smartphone software experience that repurposes old Google Pixel phones, using AI to handle virtually every task on the device.
Why it matters: First AI-branded discovery feature from the world's largest streamer signals AI entering mainstream TV recommendations.
Netflix's TV app gets an AI Discover tab showing experimental AI-powered discovery features. The first surfaces themed picks like "Something Light," "Family Movie Night" and "Quick Watches" to help viewers find something to watch.
Why it matters: Huge inference valuation signals generative media compute demand is a major funding magnet.
Fal, which hosts image, video and audio models for creators and developers, doubled its valuation to $16 billion in a $1.3 billion round led by existing investor Bessemer Venture Partners.
Iboss announced AI DLP on November 3, 2026, a data-loss prevention product inspecting content in real time across AI tools and cloud apps. A private preview opens in December 2026.
Vertafore's Velocity Change Request Agent is now available in Surefyre for managing general agents, handling change requests; internal testing cut review and documentation time 70% and errors 98%.
PrimeBOT is opening its Q1 Explorer home robot's hardware and software to owners, letting them customize the customizable home robot.
STMicroelectronics introduced a unified STM32 microcontroller platform and three STM32A product lines for high-safety automotive, industrial and robotics designs, extending a shared edge-AI foundation.
Why it matters: A 2nm AI chip marks a process-node milestone for AI silicon.
Broadcom has shipped a landmark AI chip built on a 2nm process, but investors gave the news a muted reaction, Startup Fortune reports.
Why it matters: Shows mixed-vendor GPU pooling to reduce dependence on a single accelerator ecosystem.
GPUWorks is an Apache-2.0 inference platform that pools Ascend, T-Head PPU, Hygon and NVIDIA GPUs in a single cluster; developers say one server ran 1,003 GPU machines.
Wowoker pairs Huawei Kirin hardware and HarmonyOS with sales agents that find leads, place AI calls and log every chat for employers, pitched as an AI work phone.
Nvidia App version 11.0.9.585 introduces a feature for distributing graphics card resources between programs.
A UK insurer has launched what is described as the first insurance policy covering humanoid robots, addressing liability and damage risks as robots enter workplaces.
Why it matters: Shows US pressure now blocking large AI compute deals between American infrastructure firms and Chinese giants.
Oracle abandoned a $7 billion AI chip lease deal with Tencent amid US government pressure and rising tension for American AI infrastructure firms doing business with Chinese tech giants.
Why it matters: Autonomous combat aircraft prototypes mark a step toward AI-piloted military drones.
Shield AI finished the first full-scale prototype of X-BAT, an autonomous combat aircraft with vertical flight and a claimed 2,000-mile range.
Why it matters: Allocation of scarce Nvidia chips shapes how fast Chinese AI labs can train models.
Chinese AI developers expect just a small fraction of the advanced Nvidia chips they ordered over coming months, as Nvidia struggles to boost production amid soaring demand.
BAE Systems is consolidating four configurations of its radar-equipped Bofors 40mm naval gun, pitched as containerized firepower against drones.
Why it matters: Large-scale human video pretraining for robot action models is a key physical AI frontier.
LiberAI launches Liber-0, a World Action Model pretrained on over 2.2 million hours of human video to help general-purpose robots learn physical interaction.
Lenovo Kaitian's 1.6L desktop pairs a Zhaoxin KX-7000 CPU with a Houmo LQ50 NPU for 160 TOPS, running Qwen3.6-35B-A3B locally at about 29 tokens per second.
Realbotix Echo is an $11,950 humanoid robot with a face and voice that also carries a monthly subscription fee.
A maker going by Max Imagination built a 70-gram remote-controlled robotic butterfly that flies by flapping its wings instead of using propellers.
Google's Gemini-powered smart home speaker is already discounted 20% shortly after launch, according to The Verge.
Boston Dynamics unveiled the second-generation Stretch robot, cutting cabling by 81% and reducing repair time to about one hour, per Interesting Engineering.
Stringman is an open-source, room-scale gripper robot that hangs from four ceiling cables and can tidy a whole room. Maker Nathaniel Nifong sells it for about $795–$800.
Peter Kuhar's $69 SidePulse Pro fits in the MacBook Pro SD slot, mounts as a filesystem and reads a text file to show when Claude Code, Codex and other agents need a human.
LiveX launched AI City, a project with Nvidia to bring physical AI into everyday spaces, pairing robotics and sensing in real-world environments.
Swedish defense firm Saab unveiled the CB90 Capability X, an optionally crewed combat boat that can fire missiles at speeds over 45 mph.
Nvidia's DLSS 5 neural rendering arrives in War Thunder mid-November, Where Winds Meet November 24, and RTX Remix, with claims of up to 400 FPS at 4K.
Whirlpool's Luxuriem refrigerators use 6th Sense AI to check freshness every 10 seconds; prices start at ₹98,000 in India.
Why it matters: World-model simulation could cut autonomous driving development costs and safety-driver testing time.
Baidu's Apollo World generates up to 100,000 km of simulated driving daily, about three years of road testing, producing real-time camera and lidar data to train robotaxis on rare hazards.
HP launched the OmniBook 5 and ProBook 4 G2iS AI-ready laptops in India, aimed at students and mobile professionals. Pricing and availability were not detailed.
Speechify 3.23.0 is out for macOS, Windows and Linux. It is an AI-powered reading and voice assistant app.
Amp's CLI is a command-line client for Amp, an AI coding agent and development environment. The new npm build can be installed globally with npm i -g @ampcode/cli.
Tencent is reportedly building a consumer AI search engine that combines AI-generated answers with web search, with a minimal interface and priority on authoritative sources, Huxiu reported.
HtAG is exposing Australian commercial property sales data to AI assistants through an MCP server, letting agents query the dataset directly.
Dify is an open-source collaborative workspace for building agentic workflows and RAG pipelines with support for many AI models and tools. Version 1.19.0 updates the self-hostable platform used to move prototypes to production.
Google's Dreambeans app creates personalized stories for users 18 and older.
A curated GitHub list of self-hosted open-source AI tools — local LLMs, chat UIs, agents, MCP servers, RAG and speech and image generation — health-checked weekly with dead projects removed automatically.
Version 1.24.0 of Chatbox, a desktop app for chatting with GPT-4 and GPT-3.5 through the OpenAI API, is out for macOS, Windows and Linux.
LiteLLM is an AI gateway that lets developers call 100+ LLM APIs in OpenAI or native format with cost tracking, guardrails and load balancing. v1.108.0 signs Docker images with cosign.
Antigravity 2.29.0 is an agent orchestration platform released for macOS, Windows and Linux.
OpenAI is running free ChatGPT training sessions for caregivers in five cities, aimed at helping them use the chatbot in caregiving work.
DoltHub said an agent-assisted effort produced over 15,000 lines of code and raised DumboDB's MongoDB parity tests from 3,010 to 11,359, adding 29 MongoDB operators to the database.
agent-autoloop is a TypeScript tool that has command-line agents take on and follow up tasks in the background, pinging you with cards only when they need a decision.
At IBM TechXchange 2026, IBM argued enterprise AI agents scale better when models are given structured context so they have less to infer, reducing errors.
The PC gaming mods repository Nexus Mods will require authors of new mods to declare whether AI was used in their work, opting for mandatory disclosure rather than an outright ban.
Eney is an AI assistant from MacPaw, the maker of CleanMyMac and Setapp. It runs on Mac and iPhone and plugs into Apple services like Mail, Calendar and Reminders.
LangSmith, LangChain's platform for observing and evaluating LLM apps and agents, now offers Bring Your Own Cloud deployment in beta on Microsoft Azure and Google Cloud, beyond AWS. BYOC keeps sensitive data and agent deployments in the customer's own cloud.
Ukraine's Mriia EdTech platform is using Google Gemini AI for teaching and school management, Beinsure reports.
Idol is a Gradio demo that converts a single image into a 3D object as Gaussian splats, using DINOv3 and FLUX.2-dev image encoders; running it requires an HF token with access to those models.
Uncracked City is a detective game where players interrogate suspects powered by large language models that can lie. It was posted on Show HN.
Apple's Xcode 27.3 includes GitHub Copilot in its built-in coding assistant, bringing Copilot suggestions to Apple's IDE.
Databricks MLflow now natively supports OpenTelemetry observability for Anthropic's Claude, letting teams trace agent behavior and spot issues like repeated source reads or missed findings.
Why it matters: Extends coding agents into full game-engine project editing with long unattended runs.
Kimi's game development solution pairs K3.1 models with Kimi Code and Kimi Work agents that edit Unity, Unreal Engine and Blender projects, including a 20-hour unattended run.
Basalt is a Blackwell-only inference engine for Qwen 3.8 Flash Next in NVFP4. Version 0.2.0 reports about 300 tokens/s prose, 600 tokens/s structured output and 10K tokens/s prefill on a 5090 plus 5060 Ti.
Gemini inside Gmail may soon get a redesign that looks more like the standalone Gemini app, according to Sammy Fans. No rollout date was given.
DeepL's AI-powered translation app ships version 26.11.13023090762 on macOS, Windows and Linux. The release notes give no further detail.
Why it matters: Opens Apple's assistant to third-party app actions, a key step for agentic use on iPhones.
Apple's iOS 27.3 beta gives Siri AI access to third-party apps through a developer preview for custom app integrations, and iPadOS 27.3 extends Siri mode to the Camera app.
OpenAI said its models now handle CAD, PCB layout and simulation work, citing use with Chip Ganassi Racing and Mazda. Human review is still required, the company said.
Keytomic released an MCP server, an AI visibility tracker and a Reddit AI agent aimed at automating SEO workflows for marketers.
Anthropic released version 4.16336.0 of its Claude AI desktop app for macOS, Windows and Linux.
Idol is a Gradio demo that converts a single image into a 3D object as Gaussian splats, using DINOv3 and FLUX.2-dev image encoders; running it requires an HF token with access to those models.
A developer generalized SPEED sampling nodes for MiniMax-H3 image, edit and video models into a drop-in KSampler replacement, while PrunaAI's P-H3-Image offers a 7x-smaller, 5x-faster 8-step variant on Hugging Face.
A Hugging Face GGUF of DeepSeek-V4.1-Flash-1017 compressed to about 1 bit per parameter, with half its 384 experts per layer removed, yielding 72.1 GB of weights for code and agent workflows.
A dataset of 4,146 Three.js 3D scene examples pairing prompts with full thinking traces and complete HTML/JavaScript output, spanning objects, interiors, architecture, landscapes and voxel environments. Scene code was generated with MiniMax M3.1 Flash.
A Kling research paper argues video generation's bottleneck is shifting from rendering to semantic control, summarizing multimodal understanding insights from successive Kling model generations.
YANAMI is a large-scale multimodal dataset and benchmark for multiturn stateful interaction, aimed at improving unified multimodal models that must track information across turns.
A new paper proposes ToolSpace, a method that synthesizes realistic tool environments for training long-horizon agents to compose MCP services into executable workflows.
A research paper proposes orchestrating a heterogeneous ecosystem of specialized models, tools and policies into a unified AI system, and studies model and policy routing for general-purpose AI.
A research paper, DreamX, uses a video diffusion model to learn action representations that stay consistent across executions of the same action but invariant to embodiment and appearance, aiming for better transfer across robots.
A paper presents VidUEU-Agent, an automated agent pipeline that synthesizes instruction data from raw video for multimodal understanding, editing and unified tasks, aiming to cut annotation cost.
A public Hugging Face repository for ML-Intern-lab/Qwen3.5-0.8B-weight-space-programming, an image-text-to-text model built on Qwen3.5-0.8B. The listing is a model repo, not an inference endpoint.
A paper proposes Reviso, an agentic video captioning framework arguing that not all visual evidence deserves equal attention, so MLLMs should weight key events over minor details.
CoworkOS is a research system for general-purpose agent coworkers that must deliver complex productivity work while monitoring and deciding over long horizons.
Dify is an open-source collaborative workspace for building agentic workflows and RAG pipelines with support for many AI models and tools. Version 1.19.0 updates the self-hostable platform used to move prototypes to production.
A curated GitHub list of self-hosted open-source AI tools — local LLMs, chat UIs, agents, MCP servers, RAG and speech and image generation — health-checked weekly with dead projects removed automatically.
LiteLLM is an AI gateway that lets developers call 100+ LLM APIs in OpenAI or native format with cost tracking, guardrails and load balancing. v1.108.0 signs Docker images with cosign.
agent-autoloop is a TypeScript tool that has command-line agents take on and follow up tasks in the background, pinging you with cards only when they need a decision.
ComfyUI is a node-based GUI, API and backend for running diffusion models locally. Version 0.44.0 is out; release notes give no further detail.
A GitHub repo provides an uncensored AI API and browser chat built on DeepSeek and GLM models, with streaming plus Python and TypeScript examples, pricing and setup guides.
Open WebUI is a user-friendly self-hosted interface for running local and API models (Ollama, OpenAI). Version 0.13.0 adds a tabbed workspace for files, terminal sessions, artifacts, previews, voice calls and comments.
An open GitHub project that lets AI assistants such as ChatGPT and Claude remotely control a computer through MCP, with security controls. It drew 102 stars in its first week.
Balanov/skein is a Go project for orchestration of agent swarms that fails closed, using epoch-fenced work units and rejection of late completions. It drew 116 stars in its first week.
A Python repo of 32 Agent Skills for Claude Code and Codex aimed at Chinese doctors and medical students, covering topic selection, PubMed search, bilingual writing, journal submission and peer-review responses.
A TypeScript MCP server and CLI that finds Brazilian companies by CNAE code and region using open CNPJ data from Receita Federal.
Claude Code 2.1.317 is a terminal coding agent installed via npm that understands codebases, edits files and runs commands; Pydantic AI 2.65.0, Kiro 1.4.86 and Factory Droid 1.2.0 also shipped.
An open-source TypeScript project that lets a company set up Claude Code and Codex once, self-hosted, with shared company files, team skills, MCP and per-person folder permissions.
dsh-see-now is a JavaScript tool that captures your screen, filters frames by how much they change and tiles several into one image, so AI can watch your screen in real time without spending tokens on static content.
Speechify 3.23.0 is out for macOS, Windows and Linux. It is an AI-powered reading and voice assistant app.
Tencent is reportedly building a consumer AI search engine that combines AI-generated answers with web search, with a minimal interface and priority on authoritative sources, Huxiu reported.
Google's Dreambeans app creates personalized stories for users 18 and older.
Version 1.24.0 of Chatbox, a desktop app for chatting with GPT-4 and GPT-3.5 through the OpenAI API, is out for macOS, Windows and Linux.
OpenAI is running free ChatGPT training sessions for caregivers in five cities, aimed at helping them use the chatbot in caregiving work.
The PC gaming mods repository Nexus Mods will require authors of new mods to declare whether AI was used in their work, opting for mandatory disclosure rather than an outright ban.
Eney is an AI assistant from MacPaw, the maker of CleanMyMac and Setapp. It runs on Mac and iPhone and plugs into Apple services like Mail, Calendar and Reminders.
Ukraine's Mriia EdTech platform is using Google Gemini AI for teaching and school management, Beinsure reports.
Uncracked City is a detective game where players interrogate suspects powered by large language models that can lie. It was posted on Show HN.
Gemini inside Gmail may soon get a redesign that looks more like the standalone Gemini app, according to Sammy Fans. No rollout date was given.
DeepL's AI-powered translation app ships version 26.11.13023090762 on macOS, Windows and Linux. The release notes give no further detail.
Apple's iOS 27.3 beta gives Siri AI access to third-party apps through a developer preview for custom app integrations, and iPadOS 27.3 extends Siri mode to the Camera app.
Anthropic released version 4.16336.0 of its Claude AI desktop app for macOS, Windows and Linux.
Agoda has deployed an assistant built with OpenAI to help users with travel search and booking tasks. The source gives no further product or availability details.
Itch.io added a filter letting users find or hide games that disclose generative AI use, a step Steam has not yet taken.
OpenAI added voice chat to ChatGPT's personal finance tool for US users, letting them interact with the assistant hands-free. The feature is part of OpenAI's push to make ChatGPT a personal assistant.
Gemini Notebook (NotebookLM), Google's app that turns complex material into summaries and study aids, released version 1.65.7 for iPhone and iPad with bug fixes and performance improvements.
Cyclon is a new smartphone software experience that repurposes old Google Pixel phones, using AI to handle virtually every task on the device.
Netflix's TV app gets an AI Discover tab showing experimental AI-powered discovery features. The first surfaces themed picks like "Something Light," "Family Movie Night" and "Quick Watches" to help viewers find something to watch.
TechTarget reports consumers are using AI tools to compare and shop for health insurance plans during ACA open enrollment, as they weigh premiums, networks and coverage options.
A guide shows how to build a live dashboard in Claude Artifacts that monitors Meta Ads spend, CTR, CPA and ROAS for marketers in one view.
Chinese AI companies and users are increasingly giving models and apps softer, more familiar and sometimes human-like identities, part of a push to make assistants feel more approachable.
Manus is a general AI agent that executes tasks rather than just generating text. Version 26.19.7 for iPhone and iPad adds Manus 2.0 Remote Control to drive a computer or cloud computer from the phone, plus Studio collaboration.
Mistral's Vibe agent app, which handles long-horizon tasks and can draft and deliver finished outcomes from one conversation, updated to version 2.17.0 on iPhone and iPad with various fixes and improvements; the CLI reached 3.1.0.
404 Media reports vibe-coded WoW addons are flooding the scene, prompting Warcraft Logs to change its API while developers deal with theft of their work.
Amp's CLI is a command-line client for Amp, an AI coding agent and development environment. The new npm build can be installed globally with npm i -g @ampcode/cli.
HtAG is exposing Australian commercial property sales data to AI assistants through an MCP server, letting agents query the dataset directly.
Antigravity 2.29.0 is an agent orchestration platform released for macOS, Windows and Linux.
DoltHub said an agent-assisted effort produced over 15,000 lines of code and raised DumboDB's MongoDB parity tests from 3,010 to 11,359, adding 29 MongoDB operators to the database.
At IBM TechXchange 2026, IBM argued enterprise AI agents scale better when models are given structured context so they have less to infer, reducing errors.
LangSmith, LangChain's platform for observing and evaluating LLM apps and agents, now offers Bring Your Own Cloud deployment in beta on Microsoft Azure and Google Cloud, beyond AWS. BYOC keeps sensitive data and agent deployments in the customer's own cloud.
Apple's Xcode 27.3 includes GitHub Copilot in its built-in coding assistant, bringing Copilot suggestions to Apple's IDE.
Databricks MLflow now natively supports OpenTelemetry observability for Anthropic's Claude, letting teams trace agent behavior and spot issues like repeated source reads or missed findings.
Kimi's game development solution pairs K3.1 models with Kimi Code and Kimi Work agents that edit Unity, Unreal Engine and Blender projects, including a 20-hour unattended run.
Basalt is a Blackwell-only inference engine for Qwen 3.8 Flash Next in NVFP4. Version 0.2.0 reports about 300 tokens/s prose, 600 tokens/s structured output and 10K tokens/s prefill on a 5090 plus 5060 Ti.
OpenAI said its models now handle CAD, PCB layout and simulation work, citing use with Chip Ganassi Racing and Mazda. Human review is still required, the company said.
Keytomic released an MCP server, an AI visibility tracker and a Reddit AI agent aimed at automating SEO workflows for marketers.
An analysis of the wave of AI coding agents and how they are changing what human developers do, from writing code to reviewing and directing agent output.
OpenAI built a new Git backend called Pando after its Codex AI agents overwhelmed existing version-control infrastructure, and added continuous security testing to support heavy agent-generated commits.
Google Developers blog post explains how cheap classifiers can route tasks across Gemini models and thinking levels to cut autonomous agent costs, as per-token prices fall but engineering bills rise.
Langflow Desktop is a low-code tool for building AI workflows visually. Version 1.13.0 is now available for macOS, Windows and Linux.
AWS Machine Learning · 16h
#
Amazon Quick combines structured and unstructured data in one experience; a walkthrough shows connecting S3 as a knowledge base, enriching a dataset and building a custom chat agent that answers business questions.
Amazon Web Services (AWS) · 16h · 2 sources
#
AWS published a guide for building ultra-low-latency voice AI applications and introduced India-based inference for Kimi K3 on Amazon Bedrock, plus posts on agentic semantic layers, token exchange and a Norton news engine.
Epic said Unreal Engine 5.9 Preview arrives next week, exercising an option left open at State of Unreal; it names AI workflows, web targeting, rendering and animation.
A Hugging Face blog post explains how the same model behaves differently under different agent harnesses, comparing CUGA, Hermes and Deep Agents, and when developers should pick CUGA.
Anthropic's Claude Agent SDK now can block actions when configured hooks fail, giving developers a guardrail against unintended agent behavior.
GitHub Copilot, GitHub's AI coding assistant, is now available inside JetBrains Air, letting developers get AI code completions and chat in that IDE.
TechTarget reports on the rise of agentic AI in networking, where autonomous agents manage and optimize network operations.
DeepSeek launched a Creator Mode intended to make it easier for users to build AI add-ons, according to a report.
Snowflake launched tooling to help enterprises deploy AI agents with trusted business context and built-in controls.