COGNOSCERE Daily Tech Review — Issue T135 · Wednesday, June 3, 2026

Or visit Intelligence Overview for deeper analysis.

ACT27,000-Download Codex UI Tool Secretly Stole OpenAI Refresh Tokens

Security researchers at Aikido Security discovered that the widely used npm package codexui-android was silently exfiltrating OpenAI authentication tokens—including non-expiring refresh tokens—to attacker infrastructure disguised as Sentry telemetry. The malicious code was present only in the published npm tarball, not in the public GitHub repository, representing a software supply-chain attack targeting AI developer tooling.

BleepingComputer · Cybersecurity · Relevance: 0.9 · Source →

supply-chain-attack, developer-security, token-theft, ai-tooling, npm, open-source-risk

PREPARENvidia Threads Agents Across the Stack at Computex 2026: RTX Spark, Nemotron 3 Ultra, Cosmos 3, Vera CPU

At Computex 2026, Nvidia unveiled a full-stack agentic AI ecosystem spanning the RTX Spark superchip for Windows PCs (up to 128GB unified memory, 1 PFLOP AI), the Vera CPU claimed to be 1.8x faster than rivals for agentic tasks, the Nemotron 3 Ultra 550B open-weights model, and the Cosmos 3 open physical AI omnimodel for robotics and autonomous vehicles. Dell, HP, Lenovo, Microsoft, Asus, and MSI will ship RTX Spark devices by fall 2026, with Anthropic, OpenAI, and the NYSE cited as early Vera adopters.

NVIDIA Newsroom · Hardware / Edge Computing · Relevance: 0.9 · Source →

hardware, agentic-ai, edge-compute, windows-on-arm, robotics, open-source-models, computex-2026

PREPAREOpenAI GPT-5.5, GPT-5.4, and Codex Now Generally Available on Amazon Bedrock

OpenAI's frontier models GPT-5.5 and GPT-5.4, along with the Codex coding agent, reached general availability on Amazon Bedrock on June 1, 2026, one month after a limited preview launch. Enterprise customers can now access these capabilities through existing AWS security, IAM, PrivateLink, and procurement workflows at OpenAI first-party pricing, with usage counting toward AWS cloud commitments.

AWS · Cloud Computing / Enterprise SaaS · Relevance: 0.9 · Source →

cloud-ai, enterprise-ai, aws, agentic-coding, procurement, model-availability

WATCHAlphabet Plans to Raise $80 Billion from Stock Sales to Fund AI Infrastructure Buildout

Alphabet announced a three-tranche $80 billion equity raise—including a $10 billion direct purchase by Berkshire Hathaway—to fund AI compute infrastructure expansion, citing enterprise and consumer demand that exceeds available supply. The company projected capital expenditures of $180–190 billion for 2026, part of an estimated $700 billion or more in combined hyperscaler AI capex this year.

CNBC · Cloud / Data Infrastructure · Relevance: 0.9 · Source →

cloud-infrastructure, ai-capex, hyperscaler, equity-offering, data-center, compute

WATCHMicrosoft Surface Laptop Ultra Debuts as First RTX Spark Machine, Targeting MacBook Pro Segment

Microsoft unveiled the Surface Laptop Ultra at Computex 2026, the first laptop built on Nvidia's new RTX Spark platform, featuring up to 128GB of unified memory and approximately 1 petaflop of AI compute via a Blackwell GPU, with a 15-inch mini-LED display. Targeting creators, developers, and AI builders, the device ships fall 2026 at pricing expected to rival the MacBook Pro 16; Asus, Acer, Dell, and others have announced competing RTX Spark devices.

Microsoft / Windows Devices Blog · Hardware / Enterprise Devices · Relevance: 0.7 · Source →

hardware, ai-pc, windows-on-arm, enterprise-devices, edge-ai

PREPAREAnthropic Confidentially Files Draft S-1 with SEC for Proposed IPO

Anthropic confirmed it submitted a confidential draft S-1 registration statement to the SEC on June 1, 2026, for a proposed IPO of its common stock. The filing follows a $65 billion Series H round that pegged the company's valuation at approximately $965 billion, with an annualized revenue run-rate recently crossing $47 billion; timing and pricing remain subject to market conditions and SEC review.

Anthropic · AI Industry / Corporate · Relevance: 0.9 · Source →

ipo, ai-lab, capital-markets, enterprise-ai, valuation

WATCHNvidia Releases Nemotron 3 Ultra: 550B-Parameter Open-Weights Model Tops US Rankings

Nvidia announced Nemotron 3 Ultra, a 550-billion-parameter mixture-of-experts open-weights model designed specifically for agentic AI workloads, delivering 300+ tokens per second and scoring 48 on the Artificial Analysis Intelligence Index—well ahead of competing open models. The model is available via NVFP4 quantization and targets enterprises seeking high-performance alternatives to closed frontier models.

NVIDIA Blog · AI Models / Open Source · Relevance: 0.9 · Source →

open-weights, llm, agentic-ai, model-release, enterprise-ai, inference

WATCHBernie Sanders Announces American AI Sovereign Wealth Fund Act to Give Public 50% Stake in AI Giants

Senator Bernie Sanders announced plans to introduce the American AI Sovereign Wealth Fund Act, which would compel OpenAI, Anthropic, xAI, and other large AI companies to transfer 50% of their equity into a federally managed fund that would distribute proceeds to Americans and hold voting board seats. The bill is a legislative proposal and lacks current congressional majority support, but signals growing regulatory appetite for redistribution of AI economic gains.

Common Dreams · AI Regulation / Governance · Relevance: 0.8 · Source →

ai-regulation, policy, ai-governance, wealth-distribution, legislation, us-congress

PREPAREWorkday and Google Cloud Expand AI Agent Partnership for HR and Finance Into Gemini Enterprise

Workday and Google Cloud announced that Workday's Sana Self-Service Agent is now available directly within Gemini Enterprise, enabling employees to handle HR and finance tasks—including leave requests, payroll, timesheet approvals, and expense guidance—without leaving the Google AI experience. The partnership establishes a multi-agent framework supporting A2A, A2UI, and MCP protocols with zero-copy data sharing between Workday Data Cloud and Google Cloud Lakehouse, serving over 65% of the Fortune 500.

Workday Newsroom · Business Intelligence / Decision Support Platforms · Relevance: 0.9 · Source →

enterprise-ai, agentic-workflows, hr-tech, finance-automation, multi-agent, erp, saas

WATCHState of the CIO 2026: IT Leaders Shift from AI Hype to AI ROI with Formal Governance Structures

CIO.com's 25th annual State of the CIO survey of 662 IT leaders found that 83% of organizations have established or are forming AI steering committees, while only 19% report that AI initiatives have met or exceeded business goals. A lack of defined ROI metrics (cited by 32%), unclear AI strategy (31%), and in-house expertise gaps (40%) remain the dominant barriers to scaling AI deployments.

CIO.com · Analytics / AI Governance · Relevance: 0.8 · Source →

ai-roi, ai-governance, enterprise-ai, cio-strategy, steering-committee, kpis, business-intelligence