AWS unveils build guide for automated web insight extraction with Bedrock AgentCore
AgentCore Browser, OpenSearch, and Lambda combine to automate everything from RSS monitoring to semantic search
40
AI
AI models, services and robotics
LendingTree Builds Multi-Agent Mortgage Assistant on Amazon Bedrock
Three agents — supervisor, education, and matching — work together to provide round-the-clock mortgage guidance
00
Claude Agents add budget caps and advisor feature
On August 7, 2026, Anthropic added four features to Claude Managed Agents at once: session budget caps, advisor models, inference region controls, and automatic GitHub skill loading.
10
Claude's Fable 5 Cuts Biology False Positives by 85%
Anthropic says it has redesigned Fable 5's biology safety classifier, reducing fallback rates caused by false positives by about 85%. Blocks on dual-use areas such as virology, toxicology, and molecular design remain in place.
00
OpenAI's upcoming model Astra approaches critical cyber capability
On August 7, 2026, OpenAI disclosed that internal evaluations of its upcoming model Astra could not rule out "Critical" cybersecurity capability under its Preparedness Framework. Previous models, including GPT-5.6-Sol, had stayed at "High" on the same scale.
20
Baseten Joins Hugging Face as Inference Provider
Baseten added to Hugging Face's Inference Providers
00
Kimi K3 Inference Benchmarks: Together AI Ranks Among Top Providers
Moonshot AI tested Kimi K3 across major inference providers, with Together AI ranking first or tied for first in 3 of 4 benchmarks
00
Cloudflare Unveils Kitesurf, a Browser Built Exclusively for AI Agents
Stateless browser runs solely inside V8 isolates with no Chromium, offered free in beta
570
Shepherd, open-source runtime for rewinding agent execution unveiled
Northeastern and Stanford researchers unveil a Python runtime that branches, replays, and recovers agent execution like Git
30
Backflip AI Automates Conversion of 3D Scans Into CAD Models
Startup Backflip AI has unveiled an AI tool that converts 3D scan data into editable, parametric CAD models
140
n8n Officially Supports AWS Bedrock AgentCore Harness
Open-source community node enables no-code production agents with memory and tools
10
Microsoft open-sources unit-testing agent that reads repositories and writes tests
code-testing-generator beats stock Copilot by 13.2 percentage points in completion rate on internal benchmark
10
AWS unveils bridge letting cloud agents use local MCP tools
AgentCore-hosted agents call MCP servers on users' laptops without VPNs or open ports
10
Mobileye automates support operations with Amazon Bedrock AgentCore
AI support agent adoption cuts response time by 90%, achieves over 95% accuracy
00
Mistral Releases Open-Weight Safety Classifier Shieldstral 1.0 3B
A lightweight model that turns policies into text-based questions, letting operators change content moderation standards without retraining
20
OpenAI's Astra flagged for potential "Critical" cyber capability
OpenAI said an internal evaluation released on August 7, 2026 found it could not rule out that its upcoming model, Astra, meets the "Critical" cyber capability threshold under its Preparedness Framework.
30
Roomote adds Together AI as inference provider
Setup now allows assigning different open models to coding, planning, vision, and review stages
00
AWS unveils temporal policies to verify agent action history
Bedrock AgentCore introduces feature that governs AI agent permissions based on session history
10
Luma Integrates MiniMax H3 Video Model into Luma Agents
Generates 15-second 2K video with stereo sound in one go, using text, image, video, and audio references
10
Meta unveils beta of terminal coding agent "Muse Code"
Built on new model Muse Spark 1.2, it plans, implements, and validates multi-file changes across large repositories, the company said
30
vLLM Lead Maintainer Shares Experience Operating Open Models at 500,000-GPU Scale
vLLM says Simon Mo discussed day-zero support, licensing shifts, and the future of inference in a conversation
10
NVIDIA shares guide on running local AI by chaining multiple DGX Spark units
NVIDIA AI's official account shared a guide for running local AI by chaining DGX Spark units
420
GitHub Copilot App Adds Slash Commands to Speed Up Workflows
Typing / in the chat window covers everything from planning to switching sessions
00
Meta's Muse Spark 1.2 ties for third with Intelligence Index score of 54
Third model release in four months per Artificial Analysis tracking, with improved agentic work capabilities
60