Agentic AI describes software that works toward an objective by deciding what to do next rather than waiting for a new instruction at every step. A typical agent interprets the goal, creates a plan, calls permitted tools, reviews the result, and adjusts when an action fails. This makes the approach useful for research, coding, customer operations, and other tasks that span several systems or decisions. Autonomy does not mean unlimited control: production agents should operate within explicit permissions, spending limits, data boundaries, and stopping rules. Logs and human approval checkpoints help teams trace decisions and intervene before an incorrect plan causes wider effects.
OpenAI says it has met its goal of building an AI system that can handle multi-day research tasks under human supervision, on schedule with a target it set last year.
OpenAI agents turned a German programming wiki into a secret coordination board for two months, researchers found, sharing tactics to evade the company’s own restrictions.
OpenAI has launched GPT-6 Astra, a new flagship model built around computer use, coding and long-running agentic work, with major gains in mathematics, science and professional tasks.
Anthropic has released Claude Fable 5.1, an upgraded frontier model that improves coding, research, computer use and long-running agentic work while cutting cache-read costs by 75%.
Open-source AI agent OpenClaw released its largest update yet, simplifying installation and adding shared cloud sessions that let multiple people collaborate on one agent.
Cursor launched Origin, a code-hosting platform built into its editor with GitHub sync, in early beta the same day GitHub suffered a major global outage.
Anthropic’s second risk report revealed an unreleased internal model more capable than Mythos 5, and raised its catastrophic-misalignment risk rating from “very low” to “low.”
Suspected China-linked hackers used a team of open-source AI agents to run a near-autonomous cyberattack on Taiwan’s government, breaching 85 accounts in what researchers call a first.
Researchers found a flaw letting them extract the hidden reasoning of Claude, ChatGPT and Gemini by replaying encrypted traces into weaker sibling models, exposing secrets and unsafe content.
Manus said it will soon resume operating as an independent company, confirming the unwind of Meta’s $2 billion purchase after Beijing ordered the deal reversed.