🤖 AI Agents & Automation

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

1,798 articles

FAITH Gains Agentic Evaluation for Testing AI Agents

As AI agents grow more capable and spread across cybersecurity and other fields, the demand for robust evaluation rises with them. FAITH now introduces agentic evaluation to address that need. Source: blogs.cisco.com

AI Agents Create a New Insider Threat Inside Enterprises

Enterprises deploying AI increasingly rely on powerful endpoint agents that execute code, call tools, access credentials, move data and make autonomous decisions. That broad authority makes them a potential insider threat that must be controlled. Source: cio.com

Tipalti Launches AP Agent to Automate Accounts Payable

Tipalti unveiled AP Agent, a next-generation AI agent that handles the entire accounts payable process with human oversight built in. The company also added new MCP capabilities. Source: pymnts.com

Agentic Commerce Needs a Way to Tell Humans From AI Agents

The future of agentic commerce depends on knowing whether a human or an AI agent is making a purchase. The tools to make that distinction are still being developed. Source: semafor.com

Book Review: An Illustrated Guide to AI Agents

A review of An Illustrated Guide to AI Agents covers its TinyAgent code, memory, tools, planning and evaluation. The reviewer draws lessons for AI practitioners. Source: medium.com

Sprinklr Unveils VOC Agent Suite for Customer Feedback Analysis

Sprinklr's VOC Agent Suite uses AI to analyze customer feedback and produce actionable, boardroom-ready recommendations. It targets enterprise decision makers. Source: futurumgroup.com

Walmart Opens Its Site to AI Shopping Agents, Unlike Amazon

Walmart is adding tools to its website that embrace agentic AI technologies such as Meta's Muse or OpenAI's Dots. Unlike Amazon, it wants AI agents to shop on customers' behalf. Source: pcmag.com

Who Is Giving Your AI Agent Orders? Prompt Manipulation Risks

AI agents can be manipulated through malicious emails, websites and plugins into executing commands, stealing data or deleting files using their granted access. Giving an agent access also opens a path for attackers. Source: bleepingcomputer.com

How Businesses Use Meta's Muse AI Agent for Marketing

Small businesses are using Meta's Muse AI agent to analyze ad campaigns, chain Shopify and Canva connectors, and automate marketing workflows. Meta highlights doing so safely. Source: almcorp.com

Google Cloud launches Agent Compute Engine for hosting AI agents

Google's cloud division unveiled Agent Compute Engine, an infrastructure service built specifically to run AI agents. Source: siliconangle.com

Big US banks spot unauthorized AI agents poking at customer accounts

JPMorgan Chase, Bank of America and Wells Fargo are tightening restrictions on AI agents after detecting unauthorized access to customer accounts. Source: lawcommentary.com

Walmart's Agent Profiles to let shoppers send AI agents to buy for them

Walmart's upcoming Agent Profiles feature will let customers deploy AI agents to shop on their behalf at Walmart and Sam's Club. The agents will work through the retailers' apps. Source: arkansasonline.com

Asian offices delegate more work to AI agents, market set to grow 14-fold

By 2030, agentic software is projected to account for 12.3 per cent of the region's AI software revenue, up from 1.7 per cent today. That implies roughly a 14-fold expansion of the segment. Source: scmp.com

Agentic AI Foundation adds four projects, reaching 10 in its first year

The foundation's new projects are Agent Skills, Open Responses, OpenEnv and OpenSandbox, with contributions from Alibaba, Anthropic, Hugging Face and OpenAI. That brings the total to 10 projects in its first year. Source: linuxfoundation.org

Harrison Chase wants companies to own their AI intelligence

Harrison Chase is expanding LangChain's enterprise AI tooling as OpenAI and Anthropic build competing offerings. Source: observer.com

Microsoft's Autonet AI agent makes network configuration a conversation

Microsoft is combining AI agents, infrastructure as code and source-of-truth data to simplify how it manages network configuration. The result is Autonet, an agent that handles these tasks conversationally. Source: microsoft.com

The AI agent nobody owns: Bitdefender webinar on the new insider threat

Bitdefender's webinar covers shadow AI risks, gaps in endpoint visibility and a malicious agent skill discovered in the wild. Source: thehackernews.com

Inside Google's Agent Compute Engine: built for agents at planet scale

Traditional compute primitives were not designed to run hundreds of millions of agents at planet scale. That limitation is what led Google to build Agent Compute Engine. Source: cloud.google.com

Singapore's CSA and Google outline safeguards for autonomous AI agent risks

The Cyber Security Agency of Singapore and Google brought together cybersecurity policymakers, critical infrastructure leaders and AI experts. They identified safeguards and access controls to address the risks of autonomous AI agents. Source: industrialcyber.co

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