All news with #agentic ai tag

🔒 Microsoft announced new controls to contain agentic AI workloads, including the Microsoft Execution Container (MXC) runtime and enhancements to the multi-agent vulnerability research system MDASH. MXC is a policy-driven sandbox for specifying and enforcing access to files, networks, credentials, and resources at runtime across Windows, Linux, and macOS. The company also highlighted Agent 365 SDK, Windows 365 for Agents, and two open-source standards—ASSERT and Agent Control Specifications—to govern agent behavior across platforms.

🔒 The rapid rise of frontier and agentic AI is creating board-level urgency that may finally unlock sustained cybersecurity funding. Industry leaders at recent conferences noted that autonomous AI systems expose operational risk, widen attack surfaces, and outpace traditional security architectures. CISOs are reframing cybersecurity as an operational enabler for safe AI adoption, pushing for investments in visibility, identity, monitoring, and AI-specific controls. Vendors and experts caution that budget requests need clear business cases tied to measurable outcomes.

🧭 Microsoft outlines a shift from isolated AI tools to a unified, enterprise-grade agent platform that runs real work. The post emphasizes a single integrated system spanning Azure, GitHub, Microsoft IQ, Foundry, Agent 365, and Microsoft 365 to build, contextualize, run, govern, and improve agents. It stresses secure-by-design governance, model choice, continuous improvement through feedback and tuning, and production-grade runtimes. The approach centers developers and enterprise context to make agents trustworthy and scalable.

🚀 Google Cloud Storage (GCS) is positioned as the preferred home for large-scale unstructured data and as a foundational component for production AI agents. This post highlights customer examples—Palo Alto Networks and Snap—using GCS as agent memory and analytics storage, and explains how the Model Context Protocol (MCP) enables secure, standardized access. Google offers two MCP server options: a fully managed Remote MCP server for easy, scalable deployments, and a self-managed Local MCP server for custom tooling and transformations, both integrated with Google Cloud security, observability, and tooling.

🧭 Microsoft highlights how AI-driven agentic workflows demand a shared data context. Microsoft Fabric is presented as a unified data and AI platform that enables developers and agents to build production-ready apps by providing consistent organizational context. New announcements include the open-source Rayfin SDK/CLI for rapid backend deployment and Azure HorizonDB (PostgreSQL-compatible) in public preview, optimized for AI workloads.

🧭 At Infosecurity Europe 2026, Bayer CISO Kevin Jones outlined a shift from checklist-based guidance to psychology-first security awareness to counter AI-enabled social engineering. The firm mandates behavior-focused training, ties AI access to role-based modules, and gates agent development behind completion. Bayer is moving SOCs toward supervised automation and updating supplier contracts and governance to enforce AI transparency and controls.

🔒 Amazon Quick now supports connecting privately hosted Model Context Protocol (MCP) servers via Amazon Virtual Private Cloud (VPC). This enables organizations to integrate proprietary MCP servers running on Amazon EC2, AWS Fargate, AWS Agentcore, or other private compute without exposing them to the public internet. During connector creation, choose your VPC and provide your MCP server URL so teams can interact with private MCPs in Quick while traffic remains routed securely through the VPC.

🚀 Amazon Bedrock now supports OpenAI GPT-5.5, GPT-5.4, and Codex for production use, offering the same AWS security, governance, and operational controls. GPT-5.5 delivers advanced capabilities for agentic coding, data analysis, and multi-step autonomous tasks on a next-generation inference engine. Codex is available via a dedicated App, CLI, and IDE integrations for Visual Studio Code, JetBrains, and Xcode, and can be configured to run through Bedrock with pricing aligned to OpenAI first-party rates.

🧭 At Infosecurity Europe 2026, OWASP will unveil the Agentic Research Council to better align fast‑moving agentic AI capabilities with security research and operational practice. Launched from the GenAI Security Project’s Agentic Security Initiative, the council will prioritize a public research pipeline, convene working groups and connect academic outputs to deployable mitigations. The initiative aims to accelerate runtime‑focused defenses against multi‑agent threats.

🔒 Palo Alto Networks has completed its acquisition of Portkey and will integrate Portkey’s AI Gateway into Prisma AIRS to provide a centralized control plane that secures and governs AI agents at scale. The integrated Prisma AIRS AI Gateway will offer unified APIs, an agent registry, semantic routing, artifact scanning, automated red teaming and runtime security to identify, authenticate and authorize agent interactions in real time. This aims to give enterprises a single enforcement point to manage agent identity, least-privilege access and consistent policies across autonomous workloads.

🤖 Google SRE describes how agentic AI augments traditional Site Reliability Engineering across the software lifecycle, from design and deployment to incident response and postmortems. The team applies AI agents for anomaly detection, playbook maintenance, alert enrichment, and automated mitigation while enforcing strong controls for security, explainability, and business continuity. Their approach pairs Gemini-based models and internal platforms with existing observability and governance practices.

🚀 The next generation of Amazon OpenSearch Serverless is now generally available, offering a fully managed search and vector engine optimized for agentic workflows. It auto-scales up to 20x faster and provisions resources in seconds, supports scale-to-zero and pay-per-usage pricing, and can reduce costs by up to 60% versus provisioning clusters for peak loads. New features include a shared storage layer that decouples compute and storage, two resource-based endpoints for simplified network connectivity, and native integrations with AI development platforms and OpenSearch Agent Skills.

📊 Cloudflare built Town Lake, a lakehouse-style unified data analytics platform, and Skipper, an AI data agent, to make its vast telemetry and logs queryable and auditable. Town Lake combines Trino, Iceberg on R2, a metadata catalog, PII scanning, access control, and ELT tooling to provide fresh, accurate, and governed data. Skipper lets non-SQL users ask natural-language questions, produce correct SQL-backed answers, and create shareable charts and dashboards.

🛡️ CIOs rushing to deploy AI agents without visibility risk major failures; experts warn that observability and governance are required. Many organizations treat agents like RPA and set-and-forget systems, but agents operate in model runtimes and need end-to-end tracing, least-privilege permissions, and human-in-the-loop checks. Vendors and cloud providers offer tools, yet governance can become a bottleneck if it’s not scalable and actionable.

⚠️ Anne Keast-Butler, director of GCHQ, urged UK businesses to treat cybersecurity as national defence during the agency's first annual lecture at Bletchley Park on May 27. She warned that rapid AI development narrows the window to stay ahead of threats and called on boardrooms to act now. GCHQ plans a machine-speed national cyber defence using agentic AI within five years while urging adoption of basic controls and quantum-resistant cryptography.

🚀 AWS announces Amazon Aurora MySQL-Compatible Edition now integrates with Kiro Powers, enabling developers to build Aurora MySQL-backed applications faster using AI agent assistance. The integration bundles curated Model Context Protocol (MCP) servers, steering files, and hooks validated by Kiro partners to provide immediate expertise in Aurora MySQL operations and schema design via natural language. Developers can execute data plane and control plane tasks conversationally, while task-specific guidance prevents information overload. The feature is available via one-click installation from the Kiro IDE and webpage across all Regions that support Aurora MySQL.

🛡️ Recent developments show AI moving from automation to autonomous cyber operations, shifting how offense and defense interact. The Anthropic Mythos Preview and related incidents illustrate models discovering and chaining vulnerabilities with limited human direction, prompting coordinated defensive responses from major vendors. Policy and procurement are adapting, and security leaders must treat AI agents as principals, invest in adaptive defenses, and reframe risk models for continuous compromise.

🚀 AWS announces general availability of Neuron 2.30.0, delivering NKI 0.4.0 with Trainium3-specific hardware capabilities, 22 new NKI Library kernels, and expanded Neuron Agentic Development skills to aid model porting and validation. The release targets ML developers optimizing training and inference or porting models to AWS Trainium and Inferentia, and it is available in Regions supporting Trn1/Trn2/Trn3 and Inf1/Inf2 instances.

🛡️ Anthropic appears to be preparing a public rollout of its restricted Mythos model, which the company warned poses major security risks by automating high-quality cyberattacks. Announced in April as an advanced frontier model, Mythos showed dramatic improvements in code reasoning and autonomy compared to Opus 4.7. References briefly appeared in Claude Code and Claude Security, suggesting a controlled preview, while Anthropic builds guardrails and works with partners through its Glasswing initiative.

🛡️ Researchers argue enterprises must stop treating AI agents as trusted components and instead secure them as untrusted systems. The paper, authored by teams from Google, UC San Diego, UW–Madison and others, distills five systems-security principles—least privilege, tamper resistance, complete mediation, secure information flow, and human risk—and maps eleven real-world agent attacks to these violations. They caution that stacking ML guardrails is insufficient and propose research directions for separating instructions from data, verifiable least-privilege policies, and information-flow controls.