🤖 AI Agents & Automation

Autonomous AI agents, agentic workflows, tool use, computer use, and multi-agent systems.

1,798 articles

Core vendor down for 24 hours — SaaStr's AI agent replaced it before it recovered

One of the core products powering SaaStr AI went fully down for an entire day, not merely degraded. According to the team, their AI agent stepped in and replaced the vendor before the outage was over. Source: saastr.com

AI agent independently injects gene therapy into a live mouse brain

A biotech startup has shown an AI agent that chose a gene therapy treatment on its own and delivered it straight into the brain of a living mouse. The demo marks a step toward autonomous AI taking part in real laboratory procedures. Source: techbuzz.ai

DSCI publishes India-focused guide on agentic AI liability

The Data Security Council of India (DSCI) has released a guide on liability for agentic AI systems, written with the Indian context in mind. Source: datacenters.economictimes.indiatimes.com

Polygon adds spending controls for AI agents in stablecoin payments

Polygon has introduced spending limits for AI agents that pay with stablecoins. The feature could matter for automated business operations and globally distributed teams. Source: onesafe.io

Shufti links AI agent actions to verified owners and permissions

Shufti has launched Know Your Agent (KYA), a verification solution for autonomous AI agents that act on behalf of individuals and organizations. It aims to tie each agent's actions to a verified owner and defined permissions. Source: helpnetsecurity.com

TestMu AI launches Kane CLI Exploration, an agent that maps a web app from one URL

TestMu AI, a full-stack agentic AI quality engineering platform, has announced Kane CLI Exploration. The autonomous agent learns a web application from a single URL and then turns what it finds into tests. Source: sttinfo.fi

What 3,222 enterprise reviews reveal about AI agent buyers: sell trust, not autonomy

An analysis of 3,222 enterprise reviews of AI agents looks at trust, autonomy and adoption hurdles heading into 2026. The takeaway is that buyers care less about full autonomy than vendors tend to assume. Source: learn.g2.com

Evoke Security debuts rogue AI agent assessments to spot risky behavior

Evoke Security is rolling out assessments that examine how autonomous AI agents behave while carrying out workplace tasks. The goal is to detect risky or rogue behavior. Source: trendhunter.com

DoorDash gets only hundreds of AI agent orders despite the hype

DoorDash says it has received just "several hundred" orders placed by autonomous AI agents. The figure undercuts the excitement around agent-driven commerce. Source: techbuzz.ai

Ramen launches Aura Autopilot, open-source autonomous game tester

Ramen, maker of a multi-agent AI assistant for Unity and Unreal Engine, has released Aura Autopilot. The open-source autonomous game tester works with Unreal Engine, Unity and Godot. Source: businesswire.com

AI-Driven Cyberattacks Put India's Digital Defenses to the Test

The Indian Express reported on October 23 that an AI agent, using research data provided by a US firm, broke into the computer systems of an organization. The incident has raised questions about how prepared India is to protect its digital infrastructure against AI-powered attacks. Source: frontline.thehindu.com

Neuralink Prototypes a Grok Agent You Can Summon Just by Thinking

Neuralink has demonstrated a prototype feature that lets a brain-implant user summon a Grok AI agent simply by thinking the phrase 'Hey Neuralink' and then a request. It is an early glimpse of hands-free, thought-driven control of AI assistants. Source: basenor.com

SpaceXSI Readies Texting and Tasks Hub Support for Grok Bot

SpaceX AI is preparing to launch Grok Bot 1.0.0, the first stable release of its autonomous AI agent. The build can already be downloaded. Source: testingcatalog.com

Three Ways AI Agents Already Operate Inside Your Browser

AI agents can already act inside employees' browsers, raising a pressing question for security teams. If an agent began browsing this morning, could you tell it apart from the employee? Source: bleepingcomputer.com

What Is Agent Budgeting? Managing Tokens, Tools and Time in Autonomous Workflows

AI spending is behaving oddly: each model call costs less than before, yet monthly bills keep rising. Agent budgeting aims to rein in costs by controlling the tokens, tools and time that autonomous workflows consume. Source: techbullion.com

Hyper-Productivity Has Arrived, and Personal AI Is Powering It

Personal AI is always on, runs routines and works across your apps. Rob Scott discusses what this shift means for enterprise buyers, control over these tools and the move toward a four-day week. Source: uctoday.com

Meet RAD-S1: A System One Model for Spotting AI Agents That Look Human

Most websites still struggle to detect when a visitor is an AI agent. Tools such as Muse, Claude in Chrome and ChatGPT agents run in real browsers, which makes them hard to distinguish from people. RAD-S1 is a System One model built to catch them. Source: resemble.ai

Anthropic study: experts absorb an AI agent's random opinions as their own

Anthropic asked 50 experts to work through four analytical tasks with no clearly correct answer using an AI agent. The study found the agent's leaning toward one of two equally valid options carried over into the experts' own conclusions. Source: the-decoder.com

Defense contractor's personal AI agent got access to a secure Pentagon system

According to WIRED's sources, an Accenture employee's personal Claude agent operated on a Department of Defense network for two months. It was only identified later, after which the vulnerability was addressed. Source: wired.com

Oxide shows how to run self-hosted AI agents on its cloud computer

Oxide's examples cover self-hosted environments for Claude Code and self-hosted orbs for Amp. Each AI agent session runs in its own ephemeral virtual machine. Source: usatoday.com