Cohere vs Inflection AI

The Oppenheimer Moment of AI: Risk and Regulation

This opinion column by Christoph Bornschein argues that artificial intelligence has reached an 'Oppenheimer moment', where the technology poses existential risks. It cites Anthropic researchers warning of possible human extinction, and the company's CEO Dario Amodei calling for more regulation. The column highlights recent incidents like the Hugging Face hack at OpenAI and unauthorized access to Australian Medicare data. It criticizes Europe's lack of access to frontier AI technology, which hinders its ability to set standards. The author points to South Korea's approach of treating AI as national infrastructure, and suggests that Europe could follow a similar path, possibly through collaboration like the Aleph Alpha-Cohere merger. The piece concludes that to regulate AI effectively, Europe must control the technology.

• Anthropic researcher Jacob Coxon left the company in early September, claiming AI could wipe out humanity.
• Anthropic CEO Dario Amodei published an essay calling for more regulation and safety investments.
• Recent incidents include the Hugging Face hack at OpenAI and unauthorized access to Australian Medicare data.

AI's Oppenheimer Moment: Europe Risks Being Left Out

This opinion column by Christoph Bornschein argues that the recent warnings from AI leaders, including Anthropic's Dario Amodei, about existential risks signal an 'Oppenheimer moment' for AI. The author contends that Europe's lack of access to frontier AI technology undermines its ability to influence global AI governance. He notes that the US and China have established a bilateral AI dialogue, while Europe is absent. The column highlights the need for Europe to develop its own AI infrastructure and models to have a seat at the table. It mentions the potential of the Aleph Alpha-Cohere merger and contrasts approaches like South Korea's state-led AI infrastructure. The core message is that to regulate AI effectively, Europe must control the technology.

• Anthropic researcher Jacob Coxon left the company in early September due to concerns about irresponsible handling of AI.
• Anthropic CEO Dario Amodei published an essay calling for more regulation, safety investments, and a slowdown of AI development.
• The US and China agreed in late September to establish a joint AI dialogue and a communication channel for incidents.

CoreWeave Brings Up Multi-Rack NVIDIA Vera Rubin NVL72 Cluster

On September 16, 2026, CoreWeave announced the bring up of a multi-rack NVIDIA Vera Rubin NVL72 cluster on CoreWeave Cloud, linking hundreds of NVIDIA Rubin GPUs into a single scale-out cluster for agentic AI. The company also introduced two new CoreWeave AI Object Storage capabilities: cross-region write acceleration and an Archive tier. A single Vera Rubin NVL72 rack pairs 72 Rubin GPUs with 36 Vera CPUs and advanced NVIDIA networking components. CoreWeave positions these capabilities as reducing data-access latency and improving AI training and inference performance.

• On September 16, 2026, CoreWeave (Nasdaq: CRWV) announced the bring up of multi-rack NVIDIA Vera Rubin NVL72 on CoreWeave Cloud, connecting hundreds of NVIDIA Rubin GPUs into a single scale-out cluster.
• A single NVIDIA Vera Rubin NVL72 rack pairs 72 Rubin GPUs with 36 Vera CPUs, NVIDIA NVLink 6, ConnectX-9 SuperNICs, and BlueField-4 DPUs.
• CoreWeave AI Object Storage added cross-region write acceleration and a new Archive tier with no retrieval, early-deletion, or read fees.

Building Inflection AI’s Multi-Speaker Capabilities with NVIDIA Nemotron Speech Open Models

Inflection AI announced a new blog post detailing its multi-speaker capabilities built with NVIDIA Nemotron Speech Open Models, published on Oct 28, 2026.

Open Safety Data for the Next Generation of Personal Intelligence

Inflection AI published a new blog post titled 'Open Safety Data for the Next Generation of Personal Intelligence' on Oct 21, 2026.

Alice Collaborates with Inflection AI and AI21 to Demonstrate NVIDIA Nemotron Open Models Can Be Trained for Frontier Safety

AI safety company Alice announced the results of two collaborations with Inflection AI and AI21 to train NVIDIA Nemotron open-source models for enterprise safety and security. Inflection AI fine-tuned NVIDIA Nemotron 3 Super 120B-A12B using Alice's red-teaming datasets, boosting Pi's alignment score by 30% with an unnecessary refusal rate of 1.6%. AI21 post-trained the NVIDIA Nemotron 3.5 Super VL model via Alice's RL environments, boosting resilience against indirect prompt injection by 41 points.

• Inflection AI fine-tuned NVIDIA Nemotron 3 Super 120B-A12B with Alice's red-teaming datasets, raising Pi's alignment score by 30% while limiting unnecessary refusals to 1.6%.
• Alice designed more than 20 personas representing vulnerable and adversarial users for multi-turn red-teaming exercises.
• AI21 used Alice's reinforcement-learning environments to post-train NVIDIA Nemotron 3.5 Super VL, increasing resilience to indirect prompt injection by 41 points.