🤖 AI Agents & Automation

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

1,798 articles

17 of 40 teams claiming to enforce AI agent permissions still share the underlying credentials

VentureBeat's September 2026 Agentic Security and Identity survey asked 102 enterprises how they secure and revoke AI agent access. Many teams that say they enforce permissions still reuse shared credentials behind the scenes. Source: venturebeat.com

OpenTable CEO says AI agent partnerships are bringing in new diners

The OpenTable CEO discusses how partnerships with AI agents are attracting new diners to the reservation platform. The source snippet contains little beyond the headline. Source: cnbc.com

Three days with Meta's Muse AI agent ended with a strange message and a deleted app

Ugo Nwune stopped using Muse after the agent sent him an unexpected message. He said the incident shook his trust in AI agents. Source: businessinsider.com

How AI is reshaping employment inside Anthropic

Anthropic examines how AI has restructured work and hiring within the company itself. As a frontier AI adopter, it could serve as a leading indicator of broader labor-market change. Source: anthropic.com

MBZUAI and Inception42 introduce Jais 3 and an Arabic speech recognition model

MBZUAI and Inception42 have unveiled Jais 3 alongside a new Arabic speech recognition model. The announcement was published on October 16, 2026. Source: hpcwire.com

Perplexity AI releases its secure agent sandbox SPACE as open source

AI search startup Perplexity AI Inc. is turning its attention to AI agent security following several high-profile incidents. As part of that shift, it is open-sourcing SPACE, a secure sandbox for running AI agents. Source: siliconangle.com

NVIDIA Nemotron 3.5 Super VL brings multimodal reasoning to specialized enterprise agents

Always-on agents are handling longer and more complex enterprise workflows, where a single request may mean reading a document and cross-checking information across sources. NVIDIA's Nemotron 3.5 Super VL is aimed at powering such specialized agents with multimodal reasoning. Source: developer.nvidia.com

IBM Research uses AI agent factories to design QEC decoder chips at 11.5x lower inference cost

IBM Research scientists have shown that AI "agent factories" can automate the architectural design of quantum error correction decoder chips. The approach, described in a research paper and technical release, cuts inference cost by a factor of 11.5. Source: quantumcomputingreport.com

Microsoft automates engineering operations with its Smart DRI Agent

Microsoft describes how an AI agent brings incident context, telemetry and troubleshooting into a single Teams workspace. The goal is to streamline the company's internal engineering operations. Source: microsoft.com

Banks explore 'bring your own AI agent' with the Personal Agent Protocol

Bank of America, BBVA, Synchrony and Wells Fargo have joined the Personal Agent Protocol. The framework lets companies set the parameters for what external AI agents are allowed to do. Source: americanbanker.com

Pokeworks deploys its in-house AI agent, Pokebot

The AI-powered Pokebot has automated so much work at the fast-casual chain's headquarters that the company has stopped hiring there, even as it continues to expand. Source: restaurantbusinessonline.com

Securonix debuts ABA to detect abnormal AI agent behavior

Securonix CEO Toby Weiss explains how Advanced Behavioral Analytics can reveal compromised AI agents that use valid credentials to look legitimate. Source: aimagazine.com

NVIDIA NeMo Gym used to make chip-design agents more reliable and efficient at tool calling

AI agents can boost productivity in complex chip design workflows, but they often produce plausible-looking yet wrong artifacts. NVIDIA NeMo Gym is used to improve their reliability and tool-calling efficiency. Source: developer.nvidia.com

Building trust into agentic AI through provenance records

As AI agents gain more autonomy, organizations need transparent, privacy-respecting logs of what these agents do and the reasoning behind their actions. Source: iapp.org

Kong unveils Context Management to help AI agents work efficiently

Kong Context Management lets agents reach more of the enterprise context they need without overwhelming the model with every piece of available data. Source: sdtimes.com

Swamp Club lands $7M seed round and launches its AI agent platform

Swamp Club raised $7 million in a seed round led by Amplify Partners and announced general availability of Swamp, its platform for building AI agents. Source: pulse2.com

What JPMorgan's AI agent warning signals for the banking sector

On JPMorgan's third-quarter earnings call, CEO Jamie Dimon voiced concern about AI agents, while CFO Jeremy Barnum discussed the associated risks. Source: finovate.com

Bloq Quantum releases Bloq Maya AI agent for enterprise quantum algorithm development

Quantum software developer Bloq Quantum has commercially launched Bloq Maya, an agentic AI assistant built to automate tasks such as circuit design. Source: quantumcomputingreport.com

Snowflake Cortex AI Gateway targets dangerous activity in AI agent workflows

Snowflake's Cortex AI Gateway can detect and block risky behavior across the full trajectory of an AI agent, not just individual prompts. The approach is aimed at securing long-running agent tasks. Source: snowflake.com

VentureBeat survey: OpenAI's agent platform has twice Anthropic's corporate users

VentureBeat Intelligence's September survey on agent adoption found that twice as many companies use OpenAI's agent orchestration platform as Anthropic's. Source: venturebeat.com