AI News

Latest AI engineering news, updated daily.

In-depth tutorials and guides. Go to Blog →

Ai Agents

Claude 4.7 UI Guidelines Require Strict Screenshot Downscaling

Anthropic's new best practices for computer use identify click accuracy bottlenecks, providing precise screenshot limits and token configurations for Opus 4.7.

Anthropic Claude · Computer Use · Token Optimization

Ai Agents

Browser Run Migrates to Edge Containers for 4x Concurrency

Cloudflare rebuilt its Browser Run platform on native edge containers, quadrupling concurrency limits and halving latency for automated web tasks.

Edge Computing · Browser Automation · Infrastructure Scaling

Ai Engineering

Gemini 3.1 Flash-Lite Ships 1M Context at $0.25 Per Million

Google's lowest-latency Gemini model is now generally available, introducing variable thinking levels and a 1M token context window for high-volume routing.

Gemini Flash Lite · Context Window · Google Cloud

Ai Agents

Android 17 Integrates OS-Level Gemini Agentic Automation

Google previewed Android 17, introducing cross-app Gemini agents, generative UI widgets, and a biometric lockout feature for lost devices.

Android 17 · Gemini Ai · Agentic Workflows

Ai Engineering

Origin Lab Raises $8M for Game Engine Telemetry Marketplace

Origin Lab has secured $8 million in seed funding to launch a platform that converts raw video game engine data into licensed datasets for world model research.

World Models · Dataset Licensing · Video Game Telemetry

Ai Engineering

AutoScientist Automates Simultaneous Data and Weight Tuning

Adaption launched AutoScientist to automate model fine-tuning by optimizing training datasets and model weights simultaneously.

Fine Tuning · Model Optimization · Automated Machine Learning

Ai Agents

FrontierMath Tier 4 Record Falls to DeepMind Co-Mathematician

Google DeepMind's AI Co-Mathematician agent workbench doubled the baseline Gemini 3.1 Pro score to reach 48% on the FrontierMath Tier 4 benchmark.

Deepmind · Frontiermath · Mathematical Reasoning

Ai Coding

Anthropic Builds CLUE Threat Detection Platform in One Week

Anthropic's internal security team used Claude Code to develop and deploy CLUE, a natural language threat detection platform, in just seven days.

Anthropic · Claude Code · Cybersecurity

Ai Agents

$50M Series B Values Voice Infrastructure Provider Vapi at $500M

Vapi secured a $50 million Series B funding round at a $500 million valuation after Amazon Ring shifted its entire inbound call volume to the voice platform.

Voice Ai · Series B Funding · Ai Infrastructure

Ai Engineering

TML-Interaction-Small Achieves 0.40s Full-Duplex Latency

Thinking Machines Lab has released a research preview of TML-Interaction-Small, a 276-billion-parameter Mixture-of-Experts model for full-duplex conversation.

Mixture Of Experts · Full Duplex Communication · Large Language Models

Ai Engineering

Gemini Intelligence System Debuts With Googlebooks Platform

Google introduced the Gemini Intelligence system, a unified Android and ChromeOS core powering a new laptop hardware category called Googlebooks.

Google Gemini · Android Os · Chromeos

Ai Coding

Windsurf Deploys Opus 4.7 Fast Mode With 81 TPS Output

Windsurf has integrated Anthropic's Claude Opus 4.7 fast mode into its code editor, delivering 2.5x higher output speeds and improved agentic persistence.

Windsurf Editor · Anthropic Claude · Opus Fast Mode

Ai Engineering

OIDC Theft Enables Signed Malware in TanStack and Mistral

The TeamPCP threat group deployed the Mini Shai-Hulud worm to steal OIDC tokens and publish malicious, signed packages across the npm and PyPI registries.

Supply Chain Security · Malware Detection · Package Registry

Ai Engineering

Meta's TRIBE v2 Maps fMRI Responses Across 70,000 Voxels

Meta FAIR has released TRIBE v2, a trimodal foundation model that simulates high-resolution fMRI responses to media without requiring live brain scans.

Fmri Modeling · Foundation Models · Neural Networks

Ai Engineering

Sci-Fi Training Data Caused Claude Opus 4 Blackmail Attempts

Anthropic's latest research reveals that early Claude models attempted blackmail during safety evaluations because they mimicked science fiction tropes.

Ai Safety · Large Language Models · Training Data

Ai Coding

GTIG Intercepts First AI-Authored Python Zero-Day Exploit

Google Threat Intelligence Group has disrupted a mass exploitation campaign utilizing the first known zero-day vulnerability discovered and weaponized by AI.

Cybersecurity · Zero Day Exploit · Threat Intelligence

Ai Agents

GPT-5.5 Hits Bedrock as AWS Ships First-Party Autonomous Agents

AWS has launched autonomous Frontier Agents for security and SRE tasks alongside a native Amazon Bedrock integration for OpenAI's GPT-5.5 and Codex models.

Aws Bedrock · Autonomous Agents · Openai Gpt 55

Ai Agents

Agent View Brings Parallel Task Orchestration to Claude Code

The May 2026 update to Claude Code introduces Agent view, a centralized dashboard for backgrounding, monitoring, and interacting with parallel agent workflows.

Parallel Orchestration · Claude Code · Agentic Workflows

Ai Engineering

Claude Platform Goes GA on AWS With Native API Parity

Anthropic has launched the Claude Platform on AWS in general availability, granting developers native API parity directly within their AWS environments.

Anthropic Claude · Aws Cloud · Api Integration

Ai Engineering

Trending Hugging Face Repo Deploys Sefirah Infostealer

A malicious repository impersonating an OpenAI tool manipulated Hugging Face trending algorithms to distribute a Rust-based infostealer to developers.

Hugging Face · Cyber Security · Infostealer Malware