AI News
Latest AI engineering news, updated daily.
Ai Engineering
Cloudflare AI Gateway Ships Granular Dollar Spend Limits
The new spend limits feature lets organizations set hard dollar caps across multiple LLM providers and automatically route traffic to cheaper fallback models.
Cloudflare Ai Gateway · Cost Management · Llm Infrastructure
Ai Coding
Cursor 3.7 Adds Visual Prompting to Design Mode
Cursor 3.7 introduces Design Mode, allowing developers to direct AI agents using spatial UI selections, drawn annotations, and voice commands.
Cursor Editor · Visual Prompting · Ai Design Mode
Ai Engineering
Google PHRM Achieves 6.09% MAPE in Passive Heart Rate Tracking
Google Research detailed a passive monitoring system that uses 8-second facial videos captured during routine smartphone unlocks to track resting heart rate.
Computer Vision · Biometric Monitoring · Health Tech
Ai Agents
iMessage Supports First Third-Party AI Agent With Poke Approval
Apple has approved Poke as the first AI agent for Messages for Business, allowing users to execute tasks and interact with an AI assistant directly in iMessage.
Imessage · Apple Messages For Business · Ai Automation
Ai Agents
IBM Brings Agentic SDLC Tool IBM Bob to 20000 Universities
IBM launched the Global AI Builders Challenge, expanding free access to its agentic development assistant IBM Bob across 20,000 academic institutions.
Ibm Bob · Agentic Workflow · Sdlc Automation
Ai Engineering
4B Nemotron 3.5 Content Safety Resolves AI Moderation Black Box
NVIDIA released Nemotron 3.5 Content Safety, a 4B-parameter multimodal guardrail model that provides auditable reasoning for enterprise AI moderation.
Multimodal Guardrails · Enterprise Ai · Content Moderation
Ai Agents
$200M Series F Values Coralogix's Agent Observability at $1.6B
Coralogix has raised $200 million to build observability infrastructure for autonomous AI agents, deploying MCP support and schema-free telemetry data lakes.
Observability · Venture Capital · Autonomous Agents
Ai Agents
EVA-Bench 2.0 Pits 12 Voice Models Against 213 Tasks
ServiceNow AI has released EVA-Bench Data 2.0, an open-source evaluation framework that tests conversational voice agents across 121 enterprise tools.
Voice Ai · Benchmarking · Conversational Ai
Ai Engineering
Google Dreambeans Curates Personal Data Into 14 Daily Cartoons
Google Labs has introduced Dreambeans, an experimental iOS and Android app that uses the Nano Banana 2 model to transform personal data into daily cartoons.
Google Labs · Generative Ai · Data Visualization
Ai Engineering
Open-Source ME-LSTM Framework Extends Flood Forecasts by 6 Days
Google Research has open-sourced the hydrology architecture behind Flood Hub, enabling local agencies to run ME-LSTM forecasting models on private data.
Open Source · Hydrology Modeling · Lstm Networks
Ai Coding
Claude Code Skills Cut Internal Token Usage by 47% at Anthropic
Anthropic’s internal deployment of structured skill directories in Claude Code reduces token overhead while formalizing hundreds of automated workflows.
Anthropic · Token Optimization · Automated Workflows
Ai Agents
Microsoft Debuts 35B MAI-Thinking-1 and Scout Autonomous Agents
Microsoft introduced seven in-house MAI models, the autonomous Scout workplace agent, and new Aion edge models at Build 2026 to reduce reliance on OpenAI.
Microsoft Build · Autonomous Agents · Scout Agent
Ai Engineering
Google Drops Vision Encoders in Gemma 4 12B Multimodal Release
Google DeepMind's new 12-billion parameter model uses a unified architecture to process text, image, and native audio directly on laptops with 16GB of RAM.
Multimodal Models · Edge Computing · Unified Architecture
Ai Engineering
NATO and 150 Global Partners Deploy Claude Mythos Preview
Anthropic is deploying its restricted Claude Mythos Preview model to 150 critical infrastructure organizations across 15 countries to secure core codebases.
Anthropic · Critical Infrastructure · Cybersecurity
Ai Agents
Malware Development Drives 67% of AI Cyber Misuse in 2026
Anthropic mapped 832 banned accounts to the MITRE ATT&CK framework, revealing a shift toward autonomous agent attack chains and lateral network movement.
Cybersecurity · Threat Intelligence · Autonomous Agents
Ai Engineering
Surface RTX Spark Dev Box Targets Local 120B AI Models
The new Surface RTX Spark Dev Box combines 20 Arm cores, a Blackwell GPU, and 128 GB of unified memory in a 100W chassis for local AI model fine-tuning.
Surface Rtx Spark · Local Llm · Arm Architecture
Ai Engineering
GPT-5.5 Instant Update Drops Canvas as Legacy Models Face Sunset
OpenAI updated its GPT-5.5 Instant model to reduce formulaic outputs while setting strict retirement dates for GPT-4.5 and o3 in the ChatGPT interface.
Large Language Models · Openai Updates · Model Retirement
Ai Engineering
Isaac GR00T Open Humanoid Platform Runs on Jetson Thor T5000
NVIDIA's new Isaac GR00T reference design integrates Unitree hardware, Sharpa hands, and a Blackwell-powered Jetson module for open humanoid research.
Humanoid Robotics · Nvidia Jetson · Isaac Gr00t
Ai Agents
Project Solara Drops Windows Kernel for Android AI Hardware
Microsoft's new Project Solara operating system abandons the Windows kernel for an Android foundation to power a new generation of headless AI agent devices.
Project Solara · Microsoft · Edge Computing
Ai Agents
Holo3.1 Brings 140ms Local Computer Use Agents to 12GB GPUs
Hcompany released Holo3.1, an open-weights agent framework that runs computer-use tasks locally with 140ms latency and 74.2% OS-World accuracy.
Local Inference · Open Weights · Computer Use