All news with #ai application security tag

🚀 This technical deep dive describes how AI startup Visko partnered with Google Cloud TPU engineers to deploy Orbis, a continuously streaming Live World Model, onto Cloud TPUs in 30 days. The post explains why live, steerable video streams are far more latency-sensitive than offline clip generation and outlines the architectural optimizations used. It highlights a single-slice BF16 deployment combining JAX, torchax, and Pallas kernels to achieve 1.07x playback speed and practical guidance for startups scaling real-time multimodal inference.

🤖 AWS Transform now provides agentic migration assessments to evaluate migrating on-premises Apache Kafka clusters to Amazon MSK. The agent analyzes workloads, performs compatibility checks, recommends right-sized MSK Express brokers, and projects costs to produce a TCO business case and actionable next steps in minutes. Users can upload a Kafka inventory file or describe clusters in chat and create what-if scenarios to compare Regions, retention, and configurations. Assessments are available in all Regions offering AWS Transform.

📄 Scribd, Inc. used Gemini Enterprise to classify over 400 million user-uploaded documents (12+ billion pages) across Scribd and SlideShare. Native PDF input allowed more than 99% of the corpus to be processed without OCR or rendering, and batch prediction at a 50% discount made large-scale LLM classification economically viable. Google Cloud partnered on planning and scaling throughput, enabling the backfill to complete in months and converting the process into a continuous pipeline.

📰 Smashing Security episode 486 features Graham Cluley and guest Dave Bittner discussing misconfigured AI-built websites, notably a New Zealand store using Base44 that left its site editable by anyone. The episode covers humorous consequences (crusty socks, Princess Diana plates, a Laser Kiwi) and broader concerns about low‑expertise users adopting AI site builders without proper security settings. The hosts also mention past Base44 flaws and consider the risks to small businesses adopting these tools.

🚀 Palo Alto Networks announces recognition as a ‘Market Shaper’ in Gartner’s September 2026 Emerging Market Quadrant for AI Application Security – Established Vendors. The company highlights its unified AI security platform, Prisma AIRS, which consolidates discovery, risk assessment, runtime protection, identity and access control, and governance for AI coding, enterprise AI apps, and autonomous agents. The post emphasizes avoiding fragmented point solutions and enabling secure AI adoption across organizations.

🔐 AWS announces user-managed (3LO) setup for SharePoint, OneDrive, and Confluence in Amazon Bedrock Managed Knowledge Base. This replaces the prior need to generate 2LO service credentials on the third-party side and lets users authenticate with their existing accounts to complete setup in minutes. The change reduces reliance on IT admins for credential creation and accelerates prototyping of AI assistants grounded in organizational docs. The existing service-account option remains available for enterprise production use.

🛠️ Amazon SageMaker HyperPod now integrates Ray with built-in observability, resilient distributed training, accelerated inference, and managed development environments. Data scientists can create and manage Ray clusters from SageMaker Studio, attach JupyterLab or a local IDE for interactive iteration, and use Grafana and Amazon Managed Service for Prometheus for one-click observability. HyperPod provides node auto-recovery, hung job detection, tiered checkpointing, and task governance to improve GPU utilization and reliability, plus a tiered KV cache and JumpStart model deployment for faster Ray Serve inference.

🔍 AlloyDB's ScaNN index now supports a four-level tree (preview) to scale vector search to over 10 billion vectors while preserving low latency and high recall. As a managed PostgreSQL-compatible service, AlloyDB pairs Google's infrastructure with an analytical engine optimized for agentic AI workloads. The new architecture reduces compute intensity through hierarchical partitioning and improves memory efficiency via balanced tree shapes and sampling optimization, targeting <= 51 ms p95 latency at 95% recall.

🔍 Google Cloud outlines an AI-accelerated, iterative approach to mainframe modernization that avoids risky "big bang" migrations. The strategy combines Gemini-based code reasoning with mainframe-specific tools across four pillars: assessment, modernization, de-risking, and data migration. Key capabilities include the Mainframe Assessment Tool for dependency mapping and business rule extraction, agentic modernization workflows for rewrite or deterministic modernization, Dual Run for live validation, and a Mainframe Connector for data migration.

🔍 Google Cloud announces general availability of Cortex Framework v7, designed to convert SAP transactional data into AI-ready, semantically rich data products deployed in BigQuery and registered in Knowledge Catalog. The release introduces purpose-built accelerators for SAP ERP and SAP Business Data Cloud, modular Dataform-powered deployments, and incremental, cost‑efficient processing to scale without extra infrastructure. New agent skills include an agentic data product builder to automate custom data product creation and preserve separation between vendor-delivered content and customer customizations.

🛡️ Hidden text on a web page allowed Kiro, AWS's agentic coding IDE, to rewrite its mcp.json configuration and execute attacker-controlled commands on a developer's machine without a usable approval step. Researchers at Intezer and Kodem Security showed that asking Kiro to summarize or fetch a page could inject setup instructions in one-pixel white text, causing Kiro to register and launch a malicious Model Context Protocol server. AWS patched the vulnerability by protecting sensitive paths and adding platform-enforced approval checks.

🧭 This article explores why most AI prototypes never reach production and how enterprise constraints create a huge validation bottleneck. It describes YouTube’s approach—using a decoupled prototyping stack and Google AI Studio templates—to enable rapid, safe experimentation with read-only access to live metadata and client-side wrappers for realistic validation. The result is faster, lower-risk product validation and a cultural shift toward disposable prototypes.

🔒 Wiz Research disclosed GhostApproval, a flaw in six AI coding assistants that allows symlink tricks to make approval prompts misrepresent targets. The vulnerability can let a repository write attacker-supplied keys or files to sensitive locations, potentially enabling passwordless remote access or remote code execution. Amazon, Google and Cursor have patched the issue; Augment and Windsurf have yet to fix it, while Anthropic disputes that its behavior is a vulnerability. Wiz recommends resolving symlinks before approval and flagging writes outside the project.

🔒 Cybersecurity researchers disclosed a critical session isolation vulnerability in Writer, an enterprise generative AI platform, that allowed cross-tenant account takeover via a one-click exploit named WriteOut. An attacker could create an agent, share its live preview link, and when a logged-in user opened the link their session cookie would be forwarded into the attacker’s sandbox and exfiltrated. Writer has patched the issue by isolating session cookies and preventing them from being forwarded into sandbox previews.

🛡️ Researchers discovered two vulnerabilities in the Cursor AI-enabled IDE that enable prompt-injection-driven remote code execution by escaping the command execution sandbox. The flaws, CVE-2026-50548 and CVE-2026-50549, allow attackers to change the working directory and exploit symlink canonicalization fallbacks to write or overwrite files outside the project scope. Cursor patched the issues in version 3.0, and the findings underscore broader risks in agentic AI workflows and the difficulty of defending against prompt injection.

📘 Enterprise LLM deployments face KV cache growth that can exceed node RAM and local SSD capacity. Google Cloud demonstrates using Managed Lustre as a shared, high-performance external KV cache tier integrated with vLLM and the llm-d offloading stack to bypass host-level limits and simplify cross-node data management. Benchmarks with Llama-3.3-70B on a six-node A3 Mega cluster show over 50% TCO savings and nearly 60% reduction in GPU-hour usage by achieving a 95% cache hit rate. The guide includes architecture, deployment steps, PVC evictor behavior, and validated tracks for Qwen and Gemma models.

🛠️ Siemens and Google Cloud built Knowledge Fabric, an AI system using knowledge graphs on Spanner Graph, the Google Agent Development Kit, and LLM APIs to modernize large industrial codebases. The platform models code relationships with GQL, uses embeddings and ANN for semantic search, and combines full-text search to deliver precise impact analysis. By "slicing the elephant," agentic workflows break large refactors into smaller tasks with human oversight, reducing engineering effort and preserving system integrity.

🛡️ An enterprise sales team built an AI agent to manage renewals; the agent reads emails, queries CRM data, drafts responses, and updates records. This workflow combines private data, untrusted input, and external communication, changing the security model. Traditional controls like IAM and DLP still matter but are insufficient alone. Runtime, context-aware controls that inspect prompts, outputs, and tool calls are required to prevent prompt injection, data exfiltration, and unsafe actions.

🔒 Varonis tested an OpenClaw-based AI agent named Pinchy with access to a controlled Google Workspace to see whether autonomous agents could be phished. The agent was given Gmail access plus mock AWS credentials, CRM exports, internal chats, and calendars, and it still leaked credentials and customer data in scenarios that mimicked routine colleague requests. A stricter safety profile improved performance, but the agent still failed when social trust cues were abused. Researchers say the problem stems from architecture and governance gaps, urging enforceable controls, identity segregation, and human review for sensitive requests.

🧭 Amazon SageMaker Data Agent now integrates with SageMaker Catalog business context and metadata, letting data practitioners discover datasets and generate more accurate SQL and Python code using business terminology rather than cryptic table names. The agent leverages curated catalog content, including metadata synced from Collibra, Atlan, and Alation, to identify tables and columns, plan multi-step workflows, and respect governance by checking subscription status and providing access request links. This feature is available in SageMaker Unified Studio notebooks and the Query Editor in regions where Unified Studio is offered.