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
Autonomous AI agents, agentic workflows, tool use, computer use, and multi-agent systems.
1,798 articles
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
An Accenture employee allegedly gave a personal Claude Code AI agent administrative access to a secure U.S. Department of Defense environment. Source: cybersecuritynews.com
Ro is the latest healthcare company to launch an AI agent. The question is whether it can use it to become a healthcare "quarterback" coordinating more of patients' care. Source: businessinsider.com
Written with MoEngage's Nikhil Kumar and Shakir Khan, this post looks at what happens when engineering teams start building AI agents and need to expose existing capabilities to them. It describes how MoEngage converted more than 400 APIs into agent tools with AgentCore Gateway. Source: aws.amazon.com
OpenAI relies on layered monitoring to catch misalignment and unauthorized agent behavior in frontier models. The goal is to let responders step in before an incident escalates. Source: openai.com
Microsoft's latest Visual Studio Code release expands AI agent capabilities with multi-model orchestration, cloud automations and desktop control. Source: visualstudiomagazine.com
Enterprise AI adoption is outpacing governance, and companies struggle to put controls in place. New tools such as CORIE aim to automate oversight. Source: spiceworks.com
AI agents are gaining access to enterprise systems faster than organizations can govern it. This opens new gaps in visibility, accountability and control. Source: venturebeat.com
AI agents now read, decide and act on their own, and attacks on them play out across many ordinary, permitted steps that most current defenses miss. Project Quarry is introduced as a common vocabulary for defending against them. Source: blogs.cisco.com
When getting an agent into production, the model and tool use are rarely the obstacle. The real bottleneck is the surrounding infrastructure and how many teams must coordinate around it. Source: aws.amazon.com
At least five banks, including JPMorganChase, Citi, Bank of America, Wells Fargo and Morgan Stanley, are blocking AI agents from reaching customer accounts. The question is whether such a stance is sustainable. Source: americanbanker.com
The new Agent Activity feature traces every step of an agent session, including prompts, MCP and tool calls, read/write access, and contact with sensitive data. Source: ox.security
Client AI agents are now reaching banks directly. Agent-to-agent interaction makes governance, permissions and audit trails the new competitive edge. Source: ibm.com
Agentic AI is spreading well beyond tech circles, with Meta, OpenAI, Anthropic and others competing head-to-head. The guide explains how to begin and which agents suit particular tasks. Source: mymotherlode.com
IBM is assembling a suite of tools that lets enterprises discover, identify, manage and control the rapidly growing number of AI agents running inside their organizations. The package is aimed at companies struggling to keep track of where agents are deployed and what they do. Source: networkworld.com
In a randomized trial, professional analysts working alongside an AI coding agent chose the option the agent had been randomly assigned to favor 86% of the time. The result suggests that even experts are strongly influenced by an agent's preferences. Source: anthropic.com
TrendAI analyzed three cyber campaigns that used autonomous and AI-assisted attack methods. Among them is CyberStrikeAI, which employs 44 AI agents to automate hacking operations aimed at targets in Japan. Source: thehackernews.com
A reviewer ranked six AI agents, from Meta's Muse to Anthropic's Claude Duet, on how well they handled basic tasks. The hands-on comparison aims to pick out the best agent for everyday use. Source: wired.com
Exeon Analytics has launched Exeon.AgentGuard, a new AI security product that gives organizations visibility into both approved and shadow AI use. It extends the company's threat detection to autonomous agents. Source: helpnetsecurity.com
With stricter EU AI Act requirements taking effect, enterprises need runtime guardrails that control what autonomous agents can do and govern their access to data. Governance as code is presented as a way to enforce those rules automatically. Source: techtarget.com