Releases

Oracle introduced large RDMA-connected NVIDIA GPU clusters for distributed model training.

Google DeepMind publishes AlphaMissense predictions for the effects of millions of human missense variants.

Anthropic disclosed its Long-Term Benefit Trust governance structure, designed to protect the public-benefit mission as the company scales.

Anthropic published its first Responsible Scaling Policy, introducing AI Safety Levels and deployment safeguards tied to model capability.

Oracle announced a voice-first generative-AI assistant for clinical documentation and workflows.

Databricks raised more than $500 million in Series I financing at a $43 billion valuation.

Anthropic and Boston Consulting Group partnered to help enterprise clients deploy Claude responsibly.

Microsoft Research releases Phi-1.5, a compact language model trained primarily on synthetic textbook-quality data.

Anthropic launched Claude Pro, a paid individual subscription with higher usage limits and priority access.

Baidu launched tools for fine-tuning and managing custom enterprise foundation models.

Baidu launched low-code tools for building enterprise generative-AI applications.

Baidu opened ERNIE Bot broadly after completing required generative-AI filings.

Google DeepMind and Google Cloud introduce SynthID for imperceptibly watermarking and identifying AI-generated images.

OpenAI launched ChatGPT Enterprise with enterprise security, privacy, administration, and higher-capacity access.

OpenAI named Scale a preferred partner for enterprise fine-tuning of GPT-3.5, combining Scale’s data engine with OpenAI models.

Hugging Face released Candle, a minimalist Rust framework for fast and portable machine learning.

Hugging Face raised $235 million in a Series D led by Salesforce Ventures at a $4.5 billion post-money valuation, with participation from major cloud and semiconductor companies.

Meta released code-specialized Llama models for generation, completion and instruction following.

ElevenLabs released Multilingual v2 for nearly 30 languages and moved its platform out of beta.

Hugging Face released IDEFICS, an 80B open-access visual-language model supporting sequences of images and text.

Meta released SeamlessM4T for multilingual speech-to-speech, speech-to-text, text-to-speech and text translation.

xAI completed Grok-0, a 33-billion-parameter research prototype that preceded the first public Grok model.

SK Telecom invested an additional $100 million in Anthropic and formed a partnership to build a multilingual model for telecommunications.

Scale provided evaluation infrastructure for the White House-supported Generative Red Team Challenge at DEF CON 31, testing frontier models with thousands of participants.

Hugging Face launched managed training and deployment offerings in AWS Marketplace, simplifying enterprise procurement and use on AWS.