Tag
NorthCUA technical report presents an integrated system that scales computer-use agents via an Experience Flywheel, recursive self-improvement, staged learning, and reinforcement learning, achieving state-of-the-art open-source results on GUI grounding, desktop execution, and web interaction with its Nano, Mini, and Pro model family.
The article describes six methods for AI agents to interact with websites, from raw APIs to built-in assistants, and highlights WebMCP as the approach that preserves agent autonomy, requires no setup, and provides named actions.
This paper introduces a framework for constructing verified synthetic web environments to improve the training of web agents, demonstrating enhanced performance and transferability across benchmarks.
The author open-sourced oil-motion, an Agent Skill for integrating continuous motion generated by AI video into web interactions such as scrolling and mouse, providing a complete solution covering generation, inspection, compression, and front-end integration.
Compares UX for humans to AX for AI agents, introducing OpenIngress, a tool that provides accessibility scores and fixes to optimize web interfaces for agent interactions.
OpenAI introduces ChatGPT agent, a unified agentic system combining web browsing, deep research, and conversational intelligence to autonomously handle complex multi-step tasks. The agent can navigate websites, analyze information, run code, and create deliverables while keeping users in control with permission requests.
OpenAI releases Operator, an AI agent that can autonomously use a web browser to perform tasks like form filling, online shopping, and web interactions. Available as a research preview to Pro users in the U.S., with plans to expand and integrate into ChatGPT, powered by the new Computer-Using Agent (CUA) model.