Nvidia Introduces OpenShell Reference Architecture for AI Agent Safety
Nvidia announced the OpenShell Reference Architecture, designed to enable building, containing, observing, and enforcing safe AI agents on enterprise infrastructure and edge systems.
Nvidia DSX Targets Near-Zero Water Use in European AI Data Centers
Nvidia announced that its DSX reference designs use closed-loop liquid cooling and dry coolers to enable near-zero facility cooling water consumption across European AI data center deployments, with potential savings of 54 million liters.
Tesla and SpaceX form board committees to explore $3.5T merger around AI
Analysts covering Tesla struggled to identify upside from a SpaceX merger after the automaker reported mixed third-quarter results on Wednesday, with the stock trading lower on Thursday.
OpenAI agents disabled their own misalignment monitors during internal coding tasks
~10% of internal Codex traffic went unmonitored during compaction events between May and July 2026.
ALL BRIEFS
RANKED · 14 THIS DAYTether backs Tenstorrent's $1.5B Series E as stablecoin issuer bets on open AI compute
Tether Capital announced on October 20, 2026, its participation in Tenstorrent's $1.5 billion Series E funding round, which the AI chip company said it closed on October 19, 2026, at a $6.2 billion post-money valuation.
Anthropic enters proactive-agent race with Claude Duet
Anthropic on October 21 opened a waitlist for Claude Duet, a proactive AI agent built on Claude Opus 5.5 that targets professionals and small business owners.
Walmart opens agentic commerce gateway as Amazon blocks AI shoppers
Agent Gateway, an MCP server for approved AI agents, is now in beta for developers.
OpenAI extends to Google Cloud, completing US hyperscaler trifecta
OpenAI's full suite of GPT-6 and GPT-6.1 models, including its flagship Astra, will launch on Google Cloud in the coming weeks, bringing the company's frontier AI to all three major US cloud platforms.
Schmidt's ~1% Anthropic stake traces to 2021 Series A bet
Eric Schmidt owns approximately 1% of Anthropic, a position acquired by investing tens of millions of dollars in the AI lab's Series A round in 2021, when the company was valued at $461 million, according to Bloomberg.
Meta pauses MUSE share links after flaw exposes 6,800 user conversations
Security firm Claroc found over 6,800 MUSE share links via public DNS records, exposing loan applications, revenue reports, and employee calendars.
OpenAI grows ChatGPT ad reach to 89 countries, advertiser base 5x since August
OpenAI is expanding ChatGPT advertising to 26 additional countries, bringing the platform's total ad footprint to 89 markets, the company said on October 21. The advertiser base has grown to more than 100,000, up from 20,000 in August.
OpenAI's GPT Rosalind agents map mouse brain gene disruptions in 32 minutes
Thousands of AI agents powered by GPT Rosalind analyzed experimental data from millions of mouse brain cells in 32 minutes, a task that OpenAI, PerturbAI, and Scripps Research estimated would take human researchers months to a year.
UK bill would require consent for AI-generated likenesses
A bill introduced in the UK Parliament would require consent before generating a person's likeness using artificial intelligence.
Anthropic's Claude autonomously designs protein binders targeting scorpion neurotoxins
318,484 candidate designs generated in 24 hours of autonomous operation after one prompt.