All news with #agentic ai tag

🧠 Doist built Ramble to capture stream-of-consciousness speech and convert it into structured tasks using Gemini Enterprise Agent Platform. The feature streams raw PCM audio directly to the model for simultaneous language detection, speech recognition, and semantic parsing, enabling proactive tool calls (addTask, editTask, deleteTask) and robust session resumption across devices. A layered, provider-agnostic streaming architecture supports future voice features and easy provider substitution. Outcome: fast, multilingual, real-time task capture that tolerates messy speech and offline interruptions.

🚀 AWS announced the Agent Toolkit for AWS, a production-ready suite of tools and guidance to help AI coding agents build on AWS with fewer errors, lower token costs, and enterprise-grade security controls. The toolkit includes validated agent skills, a fully managed MCP Server, and easy-to-install agent plugins. At launch AWS is shipping more than 40 skills across infrastructure-as-code, storage, analytics, serverless, containers, and AI, plus three plugin bundles for Core applications, Data Analytics, and Agent development. The Agent Toolkit is available at no additional charge; customers pay only for AWS resources consumed.

🤖 At Google Cloud Next '26, public sector leaders and academics demonstrated how the agentic era is moving from experimentation to enterprise-scale adoption across government, transportation, healthcare, and research. Featured speakers — including leaders from Google Public Sector, the City of Los Angeles, the FDA, and the Department of Transportation — shared blueprints for scaling AI and treating agents as force multipliers to improve productivity and mission outcomes. Hands-on demos, 28 Mission Talks and an interactive Public Sector Hub enabled attendees to create and test hundreds of agents across diverse use cases. Google invited organizations to continue engagement through follow-up webinars and partner pathways to accelerate adoption.

🤖 Amazon WorkSpaces now lets AI agents securely access and operate desktop applications within managed, enterprise-grade WorkSpaces environments. Agents built on any framework and running in cloud, on-premises, or hybrid deployments can connect with minimal code using the industry-standard MCP integration, while IT retains centralized permissions, logging, and auditing identical to human desktops. Observability includes screenshots and metrics for full visibility, and pay-as-you-go pricing supports elastic scale.

🔒 Google announced Agent Gateway, part of the Gemini Enterprise Agent Platform, to provide a programmable, secure connectivity plane for user-to-agent, agent-to-agent, and agent-to-tools interactions. The Gateway enables teams to inject custom logic and third-party security controls directly into the request path without changing application code. Google highlighted integrations with vendors such as Broadcom (Symantec DLP), Check Point, Cisco, CrowdStrike, Palo Alto Networks, and others to deliver runtime DLP, prompt-injection mitigation, identity governance, and behavioral analytics.

🤖 At Google Cloud Next '26, Google introduced the Gemini Enterprise Agent Platform to help teams build, deploy, scale, govern, and optimize autonomous AI agents in production. The series highlights long-running state management in Agent Runtime, a layered Agent Governance Stack, orchestration patterns in the Agent Development Kit (ADK), integration standards, and prebuilt blueprints in Agent Garden.

🤖 Join hands-on developer workshops across North America that teach secure, scalable deployment of agentic AI for enterprises. These sessions are practical, bring-your-laptop labs where Platform, Security, and Data practitioners build end-to-end solutions, including GKE cluster hardening, secure sandboxing, and governed data pipelines. Tracks cover GKE + Data and Data Engineering & Analytics, with guidance from Google experts. Attendees leave with runnable labs and operational best practices to accelerate production adoption.

🤖 Gemma 4, released by Google DeepMind, is a new family of open models optimized for local and mobile deployment. The family emphasizes intelligence per parameter, offering ultra-mobile E2B/E4B sizes, a 31B dense model for local GPUs, and a 26B Mixture-of-Experts variant. The shift to an Apache 2 license plus tools like the Agent Development Kit enables offline agentic workflows and commercial use by developers and startups.

🚀 At Google Cloud Next '26, Firestore introduced expanded agentic AI integrations, built-in full-text search, and deeper MongoDB compatibility aimed at accelerating agent-driven applications. The Enterprise edition’s reimagined query engine adds hundreds of expressive query features, JOINs via subqueries, and pipeline operations. Native connections to AI Studio and third-party coding agents plus preview tools like natural language console querying and Usage Insights simplify building agentic workflows. These capabilities are available now in Firestore Enterprise in both Native and MongoDB compatibility modes.

🤖 Amazon SageMaker AI introduces an agentic experience that dramatically shortens model customization from months to days or hours. Using SageMaker AI model customization agent skills, developers interact via natural language coding agents to prepare data, fine-tune models, evaluate quality with LLM-as-a-judge metrics, and generate reusable code artifacts. Skills can be installed into IDEs via the sagemaker-ai agent plugin or used pre-installed in SageMaker Studio Notebooks, and support deployment to Amazon Bedrock or SageMaker AI endpoints.

🔒 A joint advisory from CISA and international partners urges organizations to treat agentic AI cautiously, enforcing strong authentication, Secure by Design principles, and staged rollouts. The guidance stresses least privilege, inventories of agent capabilities, and protections against prompt injection and data exposure. It also recommends continuous monitoring with human-in-the-loop controls, DevSecOps practices, and regular incident-response testing to reduce privilege creep, tool misuse, and other emergent risks.

🔒 Microsoft announces Agent 365 is generally available, offering a unified control plane to observe, govern, and secure AI agents across endpoints, cloud, and SaaS. The release adds discovery of local and cloud agents (including OpenClaw, GitHub Copilot CLI, and Claude Code) and integrates with Intune and Defender for inventory, policy controls, runtime blocking, and alerting. Agent 365 also introduces Windows 365 for Agents, partner integrations, and licensing via Microsoft 365 E7 or standalone at USD 15 per user per month.

🛡️ CISA, in collaboration with the Australian Signals Directorate’s Australian Cyber Security Centre (ASD’s ACSC) and other partners, released guidance to help organizations adopt agentic AI systems safely. The guide identifies key security challenges and risks and offers actionable steps for designing, deploying, and operating these systems. It emphasizes risk management, alignment with existing cybersecurity frameworks, and strengthened oversight to help security teams, developers, and decision-makers implement practical governance and controls.

🔒 CISA, the Australian Signals Directorate’s Australian Cyber Security Centre (ASD ACSC), and U.S. and international partners published Careful Adoption of Agentic Artificial Intelligence Services, a joint guide describing cybersecurity challenges and mitigations for agentic AI. The document warns that agentic AI can expand attack surface, cause privilege creep, produce behavioral misalignment, and obscure event records while offering automation benefits to critical infrastructure and defense sectors. It targets developers, vendors, and operators with actionable recommendations — including avoiding broad or unrestricted access to sensitive data and systems, beginning with low‑risk, non‑sensitive use cases, and explicitly accounting for agentic AI in organizational security models and risk posture.

🔒 Microsoft previewed new Microsoft Defender capabilities within the Agent 365 tooling gateway to give security teams near real-time visibility and control over agentic workflows, using webhook-based evaluation to detect, block, and investigate anomalous agent actions before execution. Separately, Microsoft Defender for Cloud now integrates with GitHub Advanced Security generally available to map code changes to production, prioritize alerts using runtime context, and enable coordinated remediation. A hands-on Microsoft Purview demo demonstrates AI-powered data security investigations across the data estate.

🔒 At Next '26 Francis deSouza framed Google Cloud's direction as preparing enterprises for an agentic future, positioning the platform as AI‑native, open, and secure. He argued that multicloud and multi‑AI are essential for resilience and highlighted AI-driven SOC automation outcomes including a 90% reduction in mitigation time and Triage and Investigation tasks collapsed from 30 minutes to 60 seconds using Gemini. The integration of Wiz was presented as a way to secure self‑hosted and multicloud AI lifecycles, while a new whitepaper and the Gemini Enterprise Agent Platform outline governance, prompt sanitization, and agent identity controls.

🔧 AWS announced Neuron Agentic Development, an open-source set of agents and skills that bring agentic coding capabilities to development on AWS Trainium and AWS Inferentia. The initial release focuses on Neuron Kernel Interface (NKI) kernel development, enabling an agentic IDE to author, debug, profile, and analyze custom kernels. Developers can request kernels from natural-language descriptions, get automated fixes for compilation errors, and receive performance reports identifying bottleneck lines of code.

🤖 Agents can now provision Cloudflare resources and complete billing through Stripe Projects, enabling end-to-end deployment without manual dashboard steps. Using a co-designed protocol, an agent can discover available services, create or link a Cloudflare account, and receive API credentials to deploy code and register domains. Stripe supplies a payment token (not raw card data) with a default $100/month cap, and human approval can be requested when needed. Any platform with signed-in users can adopt the same orchestration flow.

🧩 Amazon Quick has introduced preview Microsoft 365 extensions for Excel, PowerPoint, and Word that let the service perform tasks directly inside users’ Microsoft 365 environments. The Excel extension supports complex spreadsheet analysis including pivot tables, charts, and data import/cleaning. PowerPoint enables template-driven deck creation and refinement from Quick data, while Word gains formatted document generation, sweeping edits with track changes, and reviewer participation in comments. The extensions are available in multiple AWS regions for early access.

⚖️ Re-permissioning aligns AI agents' access with clear operational needs so they execute tasks safely rather than accumulate unnecessary powers that enable unauthorized actions. As agents evolve from responders into execution engines, interoperability standards like MCP and agent-to-agent flows expand reach but also multiply where things can go wrong. Organizations should enforce continuous permission audits, mandatory human-in-the-loop checks for sensitive operations, strict least-privilege context sharing, and vet integrations, libraries and third parties while running tabletop prompt-injection exercises to validate controls and prevent data exposure or integrity-impacting changes.