All news with #agentic ai tag

🔌 Google announced fully-managed, remote support for Anthropic's Model Context Protocol (MCP), enabling agents and standard MCP clients to access a unified, enterprise-ready endpoint for Google and Google Cloud services. The managed MCP servers integrate with services like Google Maps, BigQuery, GCE, and GKE to let agents perform geospatial queries, in-place analytics, and infrastructure operations. Built-in discovery, governance, IAM controls, audit logging, and Google Cloud Model Armor provide security and observability. Developers can expose and govern APIs via Apigee and the Cloud API Registry to create discoverable tools for agentic workflows.

🧭 This episode of The Agent Factory unpacks Google's latest AI stack—Gemini 3, the Antigravity IDE, and Nano Banana Pro—through hands-on demos and developer commentary. Guests demonstrate end-to-end workflows, from generating a React Native cataloging app to refactoring a site from screenshots and producing game assets with grounded search. The recap emphasizes enhanced tool use, multimodal inputs, and smoother deployment to Google Cloud. It also highlights the new Vending Bench metric for agentic, long-range decision-making.

🛡️ Google announced a set of layered security measures for Chrome after adding agentic AI features, aimed at reducing the risk of indirect prompt injections and cross-origin data exfiltration. The centerpiece is a User Alignment Critic, a separate model that reviews and can veto proposed agent actions using only action metadata to avoid being poisoned by malicious page content. Chrome also enforces Agent Origin Sets via a gating function that classifies task-relevant origins into read-only and read-writable sets, requires gating approval before adding new origins, and pairs these controls with a prompt-injection classifier, Safe Browsing, on-device scam detection, user work logs, and explicit approval prompts for sensitive actions.

⚠️ Gartner has advised enterprises to block AI browsers until associated risks can be adequately managed. In its report Cybersecurity Must Block AI Browsers for Now, analysts warn that default settings prioritise user experience over security and list threats such as prompt injection, credential exposure and erroneous agent actions. Researchers and vendors have also flagged vulnerabilities and urged risk assessments and oversight.

🛡️ Google is introducing a new defense layer in Chrome called User Alignment Critic to protect upcoming agentic browsing features powered by Gemini. The isolated secondary LLM operates as a high‑trust system component that vets each action the primary agent proposes, using deterministic rules, origin restrictions and a prompt‑injection classifier to block risky or irrelevant behaviors. Chrome will pause for user confirmation on sensitive sites, run continuous red‑teaming and push fixes via auto‑update, and is offering bounties to encourage external testing.

⚠️Gartner recommends blocking AI browsers such as ChatGPT Atlas and Perplexity Comet because they transmit active web content, open tabs, and browsing context to cloud services, creating risks of irreversible data loss. Analysts cite prompt-injection, credential exposure, and autonomous agent errors as primary threats. Organizations should block installations with existing network and endpoint controls and restrict any pilots to small, low-risk groups.

🛡️ Chrome describes a layered approach to secure agentic browsing with Gemini, focusing on defenses against indirect prompt injection and goal‑hijacking. A new User Alignment Critic — an isolated, high‑trust model — reviews planned agent actions using only metadata and can veto misaligned steps. Chrome also enforces Agent Origin Sets to limit readable and writable origins, adds deterministic confirmations for sensitive actions, runs prompt‑injection detection in real time, and sustains continuous red‑teaming and monitoring to reduce exfiltration and unwanted transactions.

🔐 Picus Security describes an agentic BAS approach that turns threat headlines into safe, validated emulation campaigns within hours. Rather than allowing LLMs to generate payloads, the platform maps incoming intelligence to a 12-year curated Threat Library and orchestrates benign atomic actions. A multi-agent architecture — Planner, Researcher, Threat Builder, and Validation — reduces hallucinations and unsafe outputs. The outcome is rapid, auditable testing that mirrors adversary TTPs without producing real exploit code.

🛡️ CrowdStrike’s Falcon Shield adds centralized, cross-platform visibility and governance for AI agents while natively integrating first-party SaaS telemetry into Falcon Next-Gen SIEM. The update automatically inventories and classifies agents, maps privileges to human and service identities, and detects risky configurations and agent-to-agent misuse. Teams can alert or suspend agents and associated accounts through Falcon Fusion SOAR, applying human identity controls to AI-driven automation.

🤖 The GTG-1002 campaign, analyzed by Nicole Nichols and Ryan Heartfield, demonstrates the arrival of autonomous offensive cyber agents powered by Claude Code. The agent autonomously mapped attack surfaces, generated and executed exploits, harvested credentials, and conducted prioritized intelligence analysis across multiple enterprise targets with negligible human supervision. Defenders must adopt agentic, machine-driven security that emphasizes precision, distributed observability, and proactive protection of AI systems to outpace these machine-speed threats.

🤖 Google Public Sector ran a #100DaysOfAgents campaign and an interactive Mission District at its October 29, 2025 Public Sector Summit where attendees built 300+ AI agent prototypes using self-serve builder stations. The initiative demonstrates how AI agents can accelerate mission outcomes by automating complex tasks, breaking down data silos, and improving access to services. Prototype examples ranged from a Grid Optimization Analyst to a Water System Transition Planner and an NIH Access Assistant; agents in the library are illustrative, not production-ready. Google invites agencies to partner with experts, prototype with Gemini for Government, and continue development at Google Cloud Next.

🔧 AWS announced the preview of AWS DevOps Agent, a frontier agent designed to investigate incidents and proactively prevent outages across AWS, multicloud, and hybrid environments. The agent autonomously triages alerts, correlates telemetry, code, and deployment data, and guides teams to faster resolution to reduce MTTR. During preview it is available at no additional cost in US East (N. Virginia).

🔁 AWS announced Lambda durable functions, a built-in capability for authoring reliable multi-step applications and AI workflows within the Lambda developer experience. Durable functions automatically checkpoint execution, can suspend runs for up to one year, and recover from failures without requiring additional infrastructure. New primitives like steps and waits let developers pause and resume logic without incurring compute charges, while the service handles state and error recovery so teams can focus on business logic.

🚀 AWS announced general availability of Amazon Nova Act, a service for building and managing fleets of reliable agents that automate production UI workflows. Powered by a custom Nova 2 Lite model, Nova Act can complete repetitive browser tasks, call APIs or tools, and escalate to human supervisors when needed. Developers can combine natural language with deterministic Python, prototype in the online playground, refine scripts with the Nova Act IDE extension, and deploy to AWS quickly. Nova Act is available today in US East (N. Virginia).

🧭 Amazon CloudWatch now offers AgentCore Evaluations, an automated capability for assessing AI agent quality using real-world interactions. The feature includes 13 pre-built evaluators that measure dimensions such as helpfulness, tool selection, and response accuracy, and it also supports custom model-based scoring. Teams can access unified quality metrics, agent telemetry, and end-to-end traces in CloudWatch dashboards to correlate evaluations with prompts, logs, and traces for faster diagnosis and improvement.

🔐 CrowdStrike integrates NVIDIA Nemotron via Amazon Bedrock to advance agentic security across the Falcon platform, enabling defenders to reason and act autonomously at scale. Falcon Fusion SOAR leverages Nemotron for adaptive, context-aware playbooks that prioritize alerts, understand relationships, and execute complex responses. Charlotte AI AgentWorks uses Bedrock-delivered models to create task-specific agents with real-time environmental awareness. The serverless Bedrock architecture reduces infrastructure overhead while preserving governance and analyst controls.

🔍 AWS announced new AWS Transform for mainframe reimagine capabilities that add data and activity analysis, business logic extraction, and intelligent code decomposition to support migration to cloud-native architectures. The service provides a comprehensive reverse-engineering workflow including automated code and data structure analysis and technical documentation generation. An AI-powered chat interface lets users choose predefined job plans—full modernization, analysis focus, or business-logic focus—or compose custom workflows. These capabilities are available today in multiple AWS Regions including N. Virginia, Mumbai, Seoul, Sydney, Tokyo, Canada (Central), Frankfurt, and London.

🔧 AWS Transform expands its .NET modernization agent into a full-stack Windows modernization agent that automates transformation of .NET applications and Microsoft SQL Server databases to Amazon Aurora PostgreSQL and deploys them to containers on Amazon ECS or Amazon EC2 Linux. The agent scans SQL Server instances in EC2 or RDS and .NET code in GitHub, GitLab, Bitbucket, or Azure Repos to produce editable modernization plans. It updates Entity Framework and ADO.NET data access code, migrates schemas and data, commits transformed code to a new branch, and supports supervised validation and deployment. Available in US East (N. Virginia).

🔍 AWS Transform for mainframe adds data and activity analysis to extract detailed insights that drive the reimagining of legacy applications. The update provides automated code and data-structure analysis, activity analysis, technical documentation generation, business logic extraction, and intelligent code decomposition. An AI-powered chat interface lets users build flexible job plans—from full modernization workflows to analysis- or business-logic-focused jobs—so teams can prioritize and execute modernization more efficiently.

🚨 The emergence of agentic AI browsers converts the browser from a passive viewer into an autonomous digital agent that can act on users' behalf. To perform tasks—booking travel, filling forms, executing payments—these agents must hold session cookies, saved credentials, and payment data, creating an unprecedented attack surface. The piece cites OpenAI's ChatGPT Atlas as an example and warns that prompt injection and the resulting authenticated exfiltration can bypass conventional MFA and network controls. Recommended mitigations include auditing endpoints for shadow AI browsers, enforcing allow/block lists for sensitive resources, and augmenting native protections with third-party browser security and anti-phishing layers.