AI News, 16 November 2026

Monday 16 November

TensorWave Signs Six-Year OpenAI Deal for AMD-Powered AI Capacity

AI-generated summary TensorWave announced a six-year agreement to supply OpenAI with 457 megawatts of dedicated AI infrastructure using AMD Instinct MI455X GPUs and Helios rack-scale systems. TensorWave expects to bring the first capacity online in the first half of 2027 and complete the full deployment by early 2028.

Also covered byPulse 2.0AI PingFathom News

Core Scientific files for 285MW data center outside Dallas, Texas

AI-generated summary Core Scientific has filed plans to develop a new data center outside Dallas, Texas, with construction on the project scheduled to run from November 2026 until September 2028. AMD is set to lease 110MW in Hunt County from Core Scientific as part of a wider leasing agreement, while Core Scientific intends to fully exit its Bitcoin mining business by the end of the year.

Updating our coordinated vulnerability disclosure policy timelines

Two-panel illustration: on the left, a black-and-white photograph of a large boulder, tinted yellow, balanced on smaller rocks; on the right, a blue line drawing of mountains beneath a yellow circle on a lavender background.

AI-generated summary Anthropic says frontier models have become much better at discovering vulnerabilities, prompting updates to its coordinated vulnerability disclosure policy timelines. Anthropic says it is shortening the disclosure window in its coordinated vulnerability disclosure policy from 90 days to 30 days for high- and critical-severity issues starting January 1, 2027.

Also covered byexplainx.aiState of AIIFAI

Midas Project Finds Right-Wing Creators Recruited to Promote Nvidia Chip Sales to China

Midas Project Finds Right-Wing Creators Recruited to Promote Nvidia Chip Sales to China

AI-generated summary The Midas Project found that right-wing social media personalities were recruited to promote the export of Nvidia AI chips to China. Nvidia denied any ties to the Israeli firms involved in the influence operation and stated that it is investigating the source of the network.

Earlier:

What is the FAA's SMART AI system? Officials are denying a link to the nationwide ground stop

AI-generated summary Government officials and Air Space Intelligence report that the new SMART artificial intelligence system is not related to the recent airline ground stop. The system centralizes multiple data streams to help experts predict problems in the sky and send recommendations to the Federal Aviation Administration command center.

Earlier:

Announcing Perplexity Frontier Research.

AI-generated summary Perplexity announced Frontier Research, committing at least 10,000 GPUs to work on multi-agent orchestration and continual learning systems. The company stated that the goal is to build an automated researcher while publishing evaluations, papers, blogs, and open-sourcing models and tooling.

Also covered byRSS.bzPerplexity (@perplexity_ai)Crypto Briefing
Earlier:

Post-Training NVIDIA Nemotron to Increase Accuracy in Your Scientific Domain

AI-generated summary NVIDIA discusses post-training its Nemotron models for scientific domains to increase accuracy. NVIDIA says supervised fine tuning improved accuracy by 12% and reinforcement learning post-training lifted it a further 17%.

Also covered byNakoda NewsData PhoenixNVIDIA AI (@NVIDIAAI)
Earlier:

Numa Raises $45M Series C to Expand Its Dealership AI Operating System

AI-generated summary Numa announced a forty-five million dollar Series C funding round on November 16, 2026, to expand its dealership artificial intelligence software. Blue Cloud Ventures led the investment, bringing Numa's total funding to ninety-six million dollars.

Also covered byTNW | Artificial-intelligenceDevCurationnews.lavx.hu

How to Add a Policy Gate to MCP Tool Calls Before They Execute: vaikora-guard-mcp Explained

AI-generated summary Data443 reports that vaikora-guard-mcp is a Python proxy published on PyPI that intercepts MCP tool calls to evaluate them against configured policies before execution. The proxy records every decision in a SHA-256 hash-chained audit log and allows developers to define rules based on tool names and parameter values.

Earlier:

Baseten open-sources UltraFast engine for DeepSeek-V4.1-Flash

Baseten open-sources UltraFast engine for DeepSeek-V4.1-Flash

AI-generated summary Baseten has open-sourced the UltraFast inference engine optimized for NVIDIA B200 and B300 GPUs to support DeepSeek-V4.1-Flash. The Rust and CUDA engine achieves decoding speeds exceeding 2000 tokens per second using speculative decoding.

Also covered byOpen Model Weights
Earlier:

Michael Burry says he won't buy Nvidia stock 'at any price' as he's expecting 'real problems' ahead

AI-generated summary Michael Burry stated that he will not purchase Nvidia stock at any price due to concerns regarding the company's future and role in the artificial intelligence boom. Elon Musk warned on X that without Jensen Huang, critical agreements could be lost because many deals relied on handshakes rather than written contracts.

Also covered byInvestor's Business Daily
Earlier: