OpenAI's updated GPT-5.5 Instant is better at shopping, complex constraints, and understanding user intent — and it's already in the API

OpenAI is moving away from models that require heavy hand-holding and toward systems that can better infer the user’s goal, ...
techtimes

Gemini Computer Use Baked Into Gemini 3.5 Flash: Screen Control Now Pairs With Search and Maps

Google announced Wednesday that computer use — the ability for an AI agent to see a screen, click, type, and navigate software without a human at the keyboard — is now a built-in tool inside Gemini ...
Microsoft

AutoJack: How a single page can RCE the host running your AI agent

Ongoing research into AI agent framework security identified an exploit chain in AutoGen Studio (AutoGen’s open-source prototyping user interface) that allows untrusted web content rendered by a ...
Ars Technica

Google publishes exploit code threatening millions of Chromium users

Google on Wednesday published exploit code for an unfixed vulnerability in its Chromium browser codebase that threatens millions of people using Chrome, Microsoft Edge, and virtually all other ...
GitHub

A script to fetch data from Fitbit servers using their API and store the data in a local influxdb database.

Fitbit is a registered trademark of Google LLC. Grafana is a registered trademark of Grafana Labs. This project is an independent, open-source tool and is not ...
Wall Street Journal

Iran Has Thousands of Missiles and Could Retrieve Launchers, U.S. Intelligence Finds

Iran still has thousands of ballistic missiles in its arsenal that it could use by retrieving launchers from underground storage areas, according to American officials familiar with U.S. intelligence ...
CNBC

Anthropic says Claude can now use your computer to finish tasks for you in AI agent push

Anthropic is trialling a feature that lets users send prompts to Claude from a smartphone. Claude will complete the task on its own on a person's computer. Anthropic's product underscores its push ...
Wall Street Journal

Tech Firms Aren’t Just Encouraging Their Workers to Use AI. They’re Enforcing It.

Anthropic’s Claude AI ran a vending machine at WSJ headquarters for several weeks. It lost hundreds of dollars, bought some crazy stuff and taught us a lot about the future of AI agents. WSJ’s Joanna ...
PC World

What’s behind the OpenClaw ban wave

PCWorld reports that Anthropic and Google are banning users who connect flat-rate Claude and Gemini accounts to OpenClaw without warning or refunds. OpenClaw’s massive token consumption degrades ...
VentureBeat

OpenAI upgrades its Responses API to support agent skills and a complete terminal shell

Credit: VentureBeat made with GPT-Image-1.5 on fal.ai Until recently, the practice of building AI agents has been a bit like training a long-distance runner with a thirty-second memory. Yes, you could ...
Ars Technica

Users flock to open source Moltbot for always-on AI, despite major risks

An open source AI assistant called Moltbot (formerly “Clawdbot”) recently crossed 69,000 stars on GitHub after a month, making it one of the fastest-growing AI projects of 2026. Created by Austrian ...