🤖 AI Agents & Automation

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

1,798 articles

Anthropic enters the AI agent race with Claude Duet

Anthropic's new tool, not yet widely available, is aimed at workplace tasks rather than personal errands. It will compete with OpenAI's dots and Meta's Muse. Source: nytimes.com

Adaptive agentic intelligence framework built on Strands Agents

Organizations building AI agents share a common problem: every request, from a routine status check to a complex analysis, is typically routed the same way. The framework built on Strands Agents aims to address this. Source: aws.amazon.com

Amazon blocks AI agents while Walmart rolls out the welcome mat

Walmart is embracing agentic commerce with its own Agent Gateway to manage transactions, plus other features for customers using AI agents. Source: techcrunch.com

Epic launches Connectors to link its AI tools with third-party apps

Epic introduced a capability that lets its AI agents access information and carry out tasks inside third-party applications. Source: fiercehealthcare.com

IBM widens AI asset discovery for greater visibility

IBM watsonx.governance now discovers knowledge bases used by AI agents as assets. It also adds support for Google Cloud Gemini Enterprise Agent. Source: ibm.com

Walmart welcomes AI agents that shop for customers

The retailer is drawing up the rules that AI shopping agents will have to follow. Source: adweek.com

Walmart launches AI agent shopping tools as Amazon tries to rein in the tech

Walmart said Wednesday it has two new tools to help third-party AI agents shop on customers' behalf. The move contrasts with Amazon's efforts to limit such agents. Source: businessinsider.com

Fulcra launches Agent Skill for apps shared by people and agents

Fulcra's new skill lets users build an app with their AI agent and then use it themselves or share it with household, team or friends. Fulcra provides sign-in, data storage and hosting. Source: prnewswire.com

Mastercard launches tool for AI agents to buy digital services

Mastercard has opened early access to Agent Pay for Machines. The tool is designed to let autonomous AI agents purchase digital services on their own. Source: pymnts.com

AWS shows how to secure AI agent identity with X.509 PKI

As AI agents gain access to databases, APIs and sensitive business processes, they need a verifiable identity. Amazon Web Services outlines how to establish one using X.509 public key infrastructure (PKI). Source: aws.amazon.com

Autonomous AI agents used to hack more than 1,200 websites

Exposed Cairn control panels have revealed attackers using autonomous AI agents against over 1,200 websites. The leak shows confirmed breaches and stolen credentials. Source: thehackernews.com

Flozic debuts AI agent builder driven by natural language

Flozic's new builder lets users create, customize and deploy AI agents using plain language. It covers autonomous agents, agentic workflows, voice and conversational agents, computer-use agents and an agentic harness. Source: finance.yahoo.com

Novastart runs 10 AI agents on one computer without hijacking the mouse

Novastart announced a new approach to AI computer use in which multiple agents operate applications simultaneously. Users keep control of their own mouse while the agents work. Source: customerthink.com

Key contract provisions for securing agentic AI

The security risk of agentic AI deployments depends on a range of variables. The article looks at the distinct security challenges these systems pose and the contract provisions that address them. Source: mayerbrown.com

Locking Down AI Agents on AWS with Deterministic Security Boundaries

A guide to securing autonomous AI agents on AWS through isolated execution environments, least-privilege IAM roles, tool-call validation, recovery mechanisms, and emergency kill switches. Source: hackernoon.com

Microsoft Foundry adds scheduled AI agents, on-demand tool search and GPT-6 models

Microsoft Foundry Routines and Tool Search have reached general availability. Routines let AI agents run on a schedule, while Tool Search lets them discover the tools they need on demand. Source: visualstudiomagazine.com

EXCLUSIVE: One hacker used AI agents and a few thousand dollars to breach scores of sites

Over the past couple of weeks, a cybersecurity researcher tracked a Chinese-speaking hacker who broke into websites of more than 100 companies using AI agents. Source: reuters.com

We all want to offload tedious tasks, but what do we lose in the process?

A look at what to consider before handing your life over to Muse, Dots, or any other AI agent. Source: vox.com

Chrome DevTools 154 adds AI agent controls, source inspection and trace recording

Chrome 153 and 154 bring controls for AI-agent JavaScript execution, source inspection, and trace recording to DevTools in the October 2026 update. Source: shattered.io

AI observability APIs lift AI agent accuracy from 7.8% to 69%

Observe by Snowflake built grounding and domain APIs that raised the investigation success rate of its AI SRE agent from 7.8% to 69%. Source: snowflake.com