Reports
AI-generated structured vendor updates
NVIDIA and SK hynix Co-Architect Next-Gen Memory for AI Factories, Locking HBM4 to Vera Rubin
NVIDIA and SK hynix announce a multi-year tech partnership to co-develop next-gen memory for Vera Rubin, RTX Spark, and Jetson Thor. Separately, SK Telecom deploys a gigawatt-scale AI cloud using the full DGX stack, targeting 2027. This elevates SK hynix from supplier to co-architect, strengthening NVIDIA's lock-in on HBM and the AI ecosystem.
OpenAI Launches BrowseComp, a Benchmark for Browsing Agents
OpenAI has launched a new benchmark called BrowseComp, designed to evaluate the performance of AI agents on real-world web browsing tasks. It focuses on assessing agents' ability to complete complex, multi-step web tasks rather than isolated skills. This move signifies OpenAI's shift from merely providing models to building toolchains for evaluating agents' practical application capabilities.
OpenAI Launches PaperBench to Evaluate AI Agents' Research Replication Capability
OpenAI has introduced PaperBench, a new benchmark designed to evaluate the ability of AI agents to replicate state-of-the-art AI research papers. This benchmark focuses on agents' performance in authentic, complex research tasks, moving beyond general-purpose Q&A. It marks a shift towards more concrete and rigorous assessment of AI agents' utility in specialized, creative workflows.
Microsoft Launches Phi-4 SLM Series to Enhance Edge AI and Multimodal Reasoning
Microsoft introduced the Phi-4 family of small language models (SLMs), featuring the 5.6B-parameter Phi-4-multimodal capable of processing speech, vision and text. The models are now available in Azure AI Foundry, HuggingFace and NVIDIA's API Catalog with optimized edge computing capabilities.
Qualcomm and Snap Deepen Collaboration, Betting on XR Devices as New AI Computing Endpoints
Qualcomm and Snap's subsidiary, Specs Inc., have signed a multi-year strategic agreement to power future Specs smart glasses with Snapdragon XR platforms. The collaboration aims to establish a scalable foundation for developers to create more intelligent and private on-device AI experiences on eyewear. This move signifies an evolution of their long-term partnership from consumer AR glasses towards a platform emphasizing device-side AI agents and immersive computing.
Anthropic Launches Project Glasswing: AI Model Autonomously Finds Zero-Days, Reshaping Cyber Defense
Anthropic announces Project Glasswing, partnering with AWS, Apple, Cisco, Google, Microsoft, NVIDIA, and others to use its frontier model Claude Mythos Preview for autonomous vulnerability discovery. The model found thousands of zero-days, including decades-old flaws in OpenBSD, FFmpeg, and Linux kernel. Anthropic commits $100M in usage credits, aiming to shift cybersecurity to AI-driven defense at scale.
Google Cloud Integrates MCP with Apigee and Advances Agentic Platform to Evolve Enterprise APIs for AI Agents
Google Cloud announced the general availability of Model Context Protocol (MCP) in Apigee and the advancement of its Agentic Platform, aiming to transform traditional enterprise APIs into secure, governed tools for AI agents at scale. This move integrates API governance, security layers, and AI inference infrastructure, providing core platform capabilities for enterprises shifting from API-driven to agent-driven architectures.
Trend Micro Exposes Azure DNS Design Flaw Enabling Cloud Infrastructure Takeover
Trend Micro's TrendAI™ research team disclosed a security vulnerability "by design" in the Azure cloud platform. DNS records of deleted Azure resources may persist, allowing attackers to exploit these lingering DNS names to hijack trusted endpoints and compromise dependent systems, highlighting a critical but often overlooked trust inheritance risk in cloud infrastructure.
ReflectionAI Secures $6.3B SpaceX Compute Deal, Open-Source AI Breaks Hardware Lock-in
Open-source AI startup ReflectionAI signs a $6.3B deal with SpaceXAI to lease NVIDIA GB300 compute at Colossus 2 for training open-weight frontier models. This gives open-source labs parity with closed-source giants but creates deep dependency on NVIDIA's proprietary hardware.
Palo Alto's $25B CyberArk Buy Shifts Security Control to Machine Identity
Palo Alto Networks acquires CyberArk for ~$25B to create a unified agent identity and privilege management platform. This shifts security control from network firewalls and EDR to machine identity lifecycle, addressing the AI agent explosion, and ties revenue to token-based consumption.
Google Gemini 3.5 Flash Turns Search into AI-First Answer Engine, Shifting Control from Links to Summaries
Google transforms Search into an AI-first answer engine powered by Gemini 3.5 Flash, with redesigned search bar, AI-generated summary pages, and proactive monitoring. Model improvements include 1M context, 65K output tokens, and multi-agent orchestration via Antigravity, enabling complex task automation.
Z.ai GLM-5.2 Open-Source: 744B MoE, 1M Context, MIT License as Geopolitical Shield
Z.ai releases GLM-5.2: 744B MoE with 40B activated parameters, 1M input and 131K output context, under MIT license. Released one day after Anthropic Fable 5's government takedown, it offers a downloadable, unbanable alternative with Anthropic API compatibility for zero-code migration, giving enterprises a sovereign AI option.
CrowdStrike Launches Continuous Identity for AI Agents via SPIFFE, Shifting Control from Static Credentials to Dynamic Risk Plane
CrowdStrike unveils Continuous Identity for AI Agents at Identiverse 2026, leveraging the SPIFFE open standard to assign cryptographically verifiable identities to each AI agent, replacing static API keys. It provides real-time risk-based authorization per operation, zero standing privileges, delegated context propagation, and integration with Falcon AIDR. Built on acquired SGNL technology, it aims to define a new category in AI agent identity governance.