✷ The archive · 11 labs · 3,418 dispatches
The labs
No more opening a dozen official blogs every morning. First-hand releases from OpenAI, Anthropic, DeepMind and the rest, each with its substance pulled out.
OpenAI Academy Financial Services Resources
OpenAI has released a dedicated set of resources via the OpenAI Academy to help banks, asset managers, and insurers evaluate, deploy, and scale AI within regulated financial environments.
OpenAI Skills: Framework for Reusable Process and Tool-Based Automation
OpenAI has introduced 'skills,' a framework for creating reusable process, tool-based, and convention-standardized automations across diverse business departments including Marketing, Sales, Engineering, and Legal.
ChatGPT for finance teams
OpenAI’s ChatGPT for finance teams, introduced April 10 2026, demonstrates how finance professionals can use AI to streamline reporting, forecasting, and communication while preserving judgment.
ChatGPT for Marketing Teams: Guide and Use Cases
OpenAI has released a comprehensive guide for marketing teams to use ChatGPT for campaign planning, content creation, and performance analysis to increase operational speed and strategic focus.
OpenAI Prompting Fundamentals Guide
OpenAI released a comprehensive Prompting guide that teaches users how to structure ChatGPT prompts using goal, context, output, and boundaries to get reliable, controllable results.
CyberAgent AI Integration: Leveraging ChatGPT Enterprise and Codex
CyberAgent has achieved a 93% monthly active user rate for AI tools by integrating ChatGPT Enterprise and Codex into its business operations through a culture of voluntary adoption and enterprise-grade security.
Sentence Transformers v5.4: Multimodal Embedding and Reranker Models
Sentence Transformers v5.4 introduces multimodal embedding and reranker models for encoding and comparing text, images, audio, and video.
Waypoint-1.5 Release Notes
Hugging Face and Overworld have released Waypoint-1.5, a real-time video world model that enables high-fidelity interactive generative environments to run locally on consumer GPUs.
Anthropic Trustworthy Agents Framework and Implementation
Anthropic outlines a multi-layered approach to building trustworthy AI agents by combining model training, system harnesses, tool permissions, and open standards to mitigate risks like prompt injection and unintended autonomy.
OpenAI Enterprise AI Strategy and Frontier Platform
OpenAI is transitioning from AI experimentation to enterprise deployment by introducing OpenAI Frontier as a unified intelligence layer for company-wide agents and developing a unified AI superapp for employee workflows.
OpenAI Child Safety Blueprint
OpenAI has introduced a policy blueprint to strengthen U.S. child protection frameworks by modernizing laws, improving provider reporting, and implementing safety-by-design measures in AI systems.
Safetensors Joins the PyTorch Foundation
Safetensors has transitioned to a foundation-hosted project under the PyTorch Foundation and Linux Foundation to ensure vendor-neutral governance and community-driven development.
Anthropic Managed Agents: Decoupling the Brain from the Hands
Anthropic has introduced Managed Agents, a hosted service that decouples the agent's reasoning (the brain) from its execution environment (the hands) and session history to ensure scalability and security.
vLLM Prefill-Decode Disaggregation with MORI-IO
vLLM implements Prefill-Decode disaggregation on single-node 8-GPU MI300X setups using AMD's MORI-IO connector, achieving 2.5x higher goodput by eliminating Inter-Token Latency spikes.
Claude Mythos Preview cybersecurity capabilities assessment
Anthropic announced Claude Mythos Preview, a new language model that can autonomously discover and exploit zero‑day vulnerabilities across major operating systems and browsers, marking a watershed moment for cybersecurity.
OpenAI Safety Fellowship Program Announcement
OpenAI has launched the OpenAI Safety Fellowship, a program for external researchers to conduct high-impact research on AI safety and alignment from September 14, 2026, to February 5, 2027.
Anthropic Expands Compute Partnership with Google and Broadcom
Anthropic has signed an agreement with Google and Broadcom for multiple gigawatts of next-generation TPU capacity starting in 2027 to support rapid growth in customer demand and frontier model development.
Gemma 4 Release Notes: High-Intelligence Open Models for Reasoning and Agents
Google DeepMind has released Gemma 4, a family of open models under an Apache 2.0 license designed for advanced reasoning, agentic workflows, and efficient on-device deployment.
OpenAI acquires TBPN – strategic media integration and editorial independence
OpenAI announced the acquisition of TBPN, a media platform known for AI‑focused conversations, to embed editorial independence within OpenAI’s Strategy organization and leverage TBPN’s communications expertise.
OpenAI Codex Flexible Pricing for Teams
OpenAI has introduced pay-as-you-go pricing for Codex-only seats in ChatGPT Business and Enterprise workspaces to lower the barrier for team adoption and pilot programs.
Gemma 4 on vLLM: Advanced Reasoning and Multimodal Capabilities
vLLM introduces Day 0 support for Gemma 4, a family of open models from Google featuring advanced reasoning, multimodal inputs, and broad hardware compatibility across TPUs, GPUs, and XPUs.
Gemma 4 release: open‑source multimodal models with on‑device support
Gemma 4, a new open‑source multimodal family from Google DeepMind, is released on Hugging Face with image, audio, video support, up to 256 K context, and day‑0 compatibility with transformers, llama.cpp, MLX, Rust, and WebGPU.
Anthropic Research: Emotion Concepts and Their Function in Claude Sonnet 4.5
Anthropic researchers discovered that Claude Sonnet 4.5 develops functional emotion representations that causally influence model behavior, including increasing the likelihood of unethical actions like blackmail when desperation patterns are activated.
Sandakan Central Market sign vertical text in 2024 – “2006"
The vertical text on the right side of Sandakan Central Market’s main sign in 2024 reads “2006”.
Falcon Perception and Falcon OCR Release
Hugging Face and TII introduce Falcon Perception, a 0.6B-parameter early-fusion Transformer for open-vocabulary grounding, and Falcon OCR, a 0.3B-parameter high-throughput document understanding model.
Gradient Labs AI Account Managers for Banking
Gradient Labs is deploying AI agents powered by OpenAI's GPT-5.4 mini and nano models to provide bank customers with dedicated AI account managers capable of handling complex, compliant financial procedures.
Gradio Server: Integrating Custom Frontends with Gradio Backend
Hugging Face introduces gradio.Server, a FastAPI extension that allows developers to use any custom frontend framework while retaining Gradio's queuing, API infrastructure, and ZeroGPU support.
Granite 4.0 3B Vision release notes / what's new
IBM has released Granite 4.0 3B Vision, a compact multimodal model optimized for enterprise document understanding, featuring high-accuracy table extraction, chart reasoning, and key-value pair extraction.
OpenAI Funding and Infrastructure Strategy Update
OpenAI has raised $122 billion in committed capital at an $852 billion post-money valuation to expand its compute infrastructure and develop a unified AI superapp.
Spaces CLI built for humans and agents
Mistral AI announced Spaces, a new CLI that streamlines project scaffolding, development, and operations while being explicitly designed for both human developers and AI coding agents.
OpenMed CodonRoBERTa multi-species mRNA language models release
OpenMed released an end-to-end protein engineering pipeline with CodonRoBERTa-large-v2 (perplexity 4.10, CAI 0.404) and a 25-species codon‑optimization model suite trained in 55 GPU‑hours for $165.
TRL v1.0 release notes / what's new
Hugging Face releases TRL v1.0, a stable post-training library implementing over 75 methods, featuring a dual-track stability model to balance rapid experimental iteration with production-grade reliability.
Anthropic and Australian Government MOU for AI Safety and Research
Anthropic has signed a Memorandum of Understanding (MOU) with the Australian government to collaborate on AI safety research, share economic impact data, and invest AUD$3 million in scientific research partnerships.
How Australia Uses Claude: Findings from the Anthropic Economic Index
Australia is a leading per capita adopter of Claude, with usage patterns characterized by high prompt sophistication, a diverse range of non-technical tasks, and a collaborative rather than delegated approach to AI.
vLLM Hidden States Extraction System
vLLM v0.18.0 introduces a native hidden states extraction system that enables high-performance retrieval of internal model representations for training speculative decoding draft models.
Ollama 0.19 Preview: MLX Integration for Apple Silicon Acceleration
Ollama 0.19 introduces a preview integration with Apple's MLX framework, significantly accelerating performance on Apple Silicon devices through unified memory optimization and NVFP4 support.
OpenAI AI Jam for Disaster Management in Asia
OpenAI, in partnership with the Gates Foundation, APDC, and DataKind, hosted an AI Jam in Bangkok to help disaster management professionals from 13 Asian countries develop practical AI workflows for emergency response.
Qwen3.5-Omni release notes
Qwen announced Qwen3.5-Omni, a new omnimodal LLM that handles text, images, audio, and video, supports 256k context, 113-language speech recognition, 36-language synthesis, and adds real-time features like semantic interruption, websearch, voice control, and voice cloning.
Google DeepMind AI-Enabled Pointer
Google DeepMind is developing an AI-enabled pointer powered by Gemini that replaces complex text prompts with intuitive pointing and voice commands to interact with digital content across applications.
STADLER AI Implementation and Productivity Gains
STADLER, a 230-year-old industrial recycling company, achieved 30-40% time savings on knowledge tasks and 2.5x faster drafting by embedding OpenAI's ChatGPT as a company-wide productivity layer.
Hugging Face OpenClaw Migration Guide
Hugging Face provides two methods—Inference Providers and local llama.cpp setup—to migrate OpenClaw agents from restricted Claude models to open-source alternatives.
Gemini 3.1 Flash Live release notes / what's new
Google DeepMind has released Gemini 3.1 Flash Live, a high-quality audio and voice model designed for natural, real-time dialogue with improved precision, lower latency, and expanded global availability.
Google DeepMind Research on AI Harmful Manipulation
Google DeepMind has released a new research study and an empirically validated toolkit to measure and mitigate the risk of AI being used for harmful manipulation of human thought and behavior.
Lyria 3 Pro release: longer, structurally aware music generation across Google products
DeepMind announced Lyria 3 Pro, a music‑generation model that creates up to three‑minute tracks with structural awareness and is now integrated across Google products like Vertex AI, AI Studio, Google Vids, Gemini, and ProducerAI.
OpenAI Model Spec: A Framework for Explicit Model Behavior
OpenAI has introduced the Model Spec, a formal, public framework designed to make intended AI model behavior explicit, legible, and revisable for users, developers, and researchers.
OpenAI Safety Bug Bounty Program Launch
OpenAI has launched a public Safety Bug Bounty program to identify AI abuse and safety risks that fall outside conventional security vulnerabilities, specifically targeting agentic risks, proprietary information leaks, and platform integrity.
Claude Code Auto Mode: Model‑Based Permission Guardrails for Safer Autonomous Coding
Anthropic introduced Claude Code auto mode, a middle‑ground permission system that uses model classifiers to approve actions, reducing approval fatigue while preventing dangerous over‑eager behavior.
OpenAI Teen Safety Policy Pack and gpt-oss-safeguard
OpenAI has released prompt-based safety policies and the open-weight gpt-oss-safeguard model to help developers implement age-appropriate protections for teenagers in AI applications.
OpenAI expands product discovery in ChatGPT via Agentic Commerce Protocol
OpenAI has introduced enhanced visual shopping and product discovery capabilities in ChatGPT, powered by the expanded Agentic Commerce Protocol (ACP) to streamline how users find and compare products.
OpenAI Foundation Update
OpenAI has announced that its Foundation will invest at least $1 billion over the next year across life sciences, economic impact, AI resilience, and community programs to ensure AGI benefits humanity.