✷ The archive · 11 labs · 3,054 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 Scam Operations Report October 2025
OpenAI has disrupted scam networks originating in Cambodia, Myanmar, and Nigeria that used ChatGPT for scaling fraud, while noting that users employ the model to identify scams three times more often than scammers use it for fraud.
OpenAI Disrupts Russian-Speaking Malware Tooling Operation
OpenAI has banned accounts affiliated with Russian-speaking criminal groups that used ChatGPT to prototype and refine malware components, including credential stealers and evasion techniques.
OpenAI Disrupts Cyber Operation Targeting Semiconductor and Academic Sectors
OpenAI banned a cluster of ChatGPT accounts used by threat actors linked to PRC intelligence requirements to enhance phishing efficiency and develop low-to-mid maturity malware.
OpenAI Disrupts Korean-language Malware Development Operation
OpenAI has banned a cluster of ChatGPT accounts used by Korean-language operators to develop malware, command-and-control infrastructure, and phishing campaigns, though no novel capabilities were enabled.
OpenAI Disrupts PRC-Linked AI Abuse for Surveillance and Influence
OpenAI has banned accounts linked to PRC government entities that used ChatGPT to design surveillance tools, develop profiling proposals, and conduct targeted online research.
OpenAI Operation Stop News Report
OpenAI banned accounts linked to a Russian-origin covert influence operation that used AI to generate text, video scripts, and SEO content for fake news outlets in Africa and the UK.
Operation Nine—emdash Line: Disrupting China-Origin Influence Activity
OpenAI has banned a network of ChatGPT accounts used by a China-affiliated covert influence operation to generate social media content targeting the South China Sea, Hong Kong, and the Philippines.
OpenAI Sora 2 System Card
OpenAI has introduced Sora 2, a state-of-the-art video and audio generation model featuring improved physics, synchronized audio, and enhanced realism.
OpenAI Sora 2 and Sora App Safety Framework
OpenAI has launched Sora 2 and the Sora app, introducing a comprehensive safety framework featuring C2PA metadata, consent-based character likeness, and specialized protections for teen users.
OpenAI Sora 2 Release
OpenAI has released Sora 2, a flagship video and audio generation model featuring improved physical accuracy, synchronized sound, and a new social iOS app for co-creative experiences.
OpenAI Research Assistant: Internal Tool for Scaling Customer Insights
OpenAI has developed an internal research assistant powered by GPT-5 to transform millions of support tickets into actionable product insights in minutes rather than weeks.
OpenAI Contract Data Agent announcement
OpenAI unveiled an internal Contract Data Agent that automates contract extraction and reasoning, halving review time and enabling finance to scale without proportional headcount growth.
OpenAI Support Model: Transforming User Support into an AI Operating Model
OpenAI has implemented an internal AI-driven support operating model that transforms support representatives into system builders who continuously improve the knowledge base and model performance through a feedback loop of surfaces, knowledge, and evals.
OpenAI on OpenAI series showcases internal AI tools that turn expertise into core business infrastructure
OpenAI announced the OpenAI on OpenAI series, revealing internal AI tools that boost productivity across sales, finance, research, and support, demonstrating how frontier AI can become core business infrastructure.
OpenAI Inbound Sales Assistant Announcement
OpenAI introduced an AI‑driven inbound sales assistant that delivers personalized, accurate answers to thousands of prospects, turning missed leads into multimillion‑dollar revenue growth.
OpenAI GTM Assistant: Scaling Sales Productivity with Internal AI Automation
OpenAI developed GTM Assistant, an internal Slack-based tool that increases sales rep productivity by 20% through automated customer research and product Q&A.
OpenAI Child Safety and CSAM Prevention Strategy
OpenAI has detailed its multi-layered approach to preventing, detecting, and reporting child sexual abuse material (CSAM) and child sexual exploitation material (CSEM) across its AI models and developer ecosystem.
OpenAI Parental Controls Release
OpenAI has introduced parental controls for ChatGPT, allowing parents to link accounts with their teens to manage safety settings, set usage limits, and receive safety notifications.
VibeGame: A High-Level Declarative Engine for AI-Assisted Game Development
Hugging Face researcher Dylan Ebert introduces VibeGame, a declarative game engine designed to enable 'vibe coding' by combining high-level abstractions with a web-based stack optimized for AI proficiency.
OpenAI Buy it in ChatGPT and Agentic Commerce Protocol
OpenAI has introduced Instant Checkout and the open-source Agentic Commerce Protocol to enable direct product purchases within ChatGPT for U.S. users.
Qwen3-8B acceleration on Intel Core Ultra with depth‑pruned draft models
Accelerated Qwen3-8B on Intel Core Ultra using speculative decoding and a depth‑pruned Qwen3‑0.6B draft achieves ~1.4× speedup, enabling fast local AI agents via Hugging Face smolagents.
Claude Code Updates: VS Code Extension, Checkpoints, and Agent SDK
Anthropic has updated Claude Code with a native VS Code extension, a checkpointing system for autonomous task management, and an expanded Agent SDK powered by Claude Sonnet 4.5.
Anthropic Effective Context Engineering for AI Agents
Anthropic introduces context engineering as a strategy for curating the optimal set of tokens to maximize LLM performance and overcome attention scarcity and context rot.
DeepSeek-V3.2-Exp Release Notes
DeepSeek has released DeepSeek-V3.2-Exp, an experimental model featuring DeepSeek Sparse Attention (DSA) to improve long-context efficiency and reduce API pricing by over 50%.
Anthropic Consumer Terms and Privacy Policy Updates
Anthropic has updated its Consumer Terms and Privacy Policy to allow users on Free, Pro, and Max plans to opt-in to using their data for model training and safety improvements, with extended data retention for those who participate.
Nemotron-Personas-Japan Release
NVIDIA has released Nemotron-Personas-Japan, an open synthetic dataset of 6 million Japanese personas designed to support the development of Sovereign AI and culturally aware LLMs.
OpenAI and AARP Partnership for Older Adult Online Safety
OpenAI has partnered with AARP and its Older Adults Technology Services (OATS) to provide AI literacy training and scam prevention tools for older adults.
Swift Transformers 1.0 release notes / what's new
Hugging Face has released Swift Transformers 1.0, a stable library providing tokenizers, hub access, and model wrappers to simplify local LLM integration for Apple Silicon developers.
Anthropic Global Enterprise Expansion and Leadership Appointments
Anthropic has appointed Chris Ciauri as Managing Director of International to lead its global expansion, supported by a revenue increase to over $5 billion run-rate as of August 2025.
ChatGPT Business Updates: Shared Projects, Tool Connectors, and Enhanced Compliance
OpenAI has introduced shared projects for ChatGPT business plans, expanded tool connectors for Gmail and Microsoft 365, and added new ISO certifications to improve team collaboration and enterprise security.
OpenAI GDPval: Measuring Model Performance on Real-World Knowledge Work
OpenAI has introduced GDPval, a new evaluation framework that measures AI model performance on 1,320 economically valuable tasks across 44 professional occupations to track real-world productivity gains.
ChatGPT Pulse Preview
OpenAI has introduced ChatGPT Pulse, a proactive research feature for Pro users on mobile that delivers personalized daily updates based on user history, feedback, and connected apps.
xAI expands government access via GSA OneGov initiative
xAI is making its frontier reasoning models, including Grok 4 and Grok 4 fast, available to all US federal government departments and agencies for $0.42 over the next 18 months as part of the GSA OneGov initiative.
ENEOS Materials Adopts ChatGPT Enterprise for Manufacturing Productivity
ENEOS Materials has integrated ChatGPT Enterprise across its organization to combat labor shortages and rising costs, achieving significant reductions in research time and data analysis workflows.
OpenAI for Germany partnership announcement
OpenAI and SAP announced OpenAI for Germany, a sovereign AI platform for the German public sector that combines OpenAI models with SAP's Delos Cloud on Azure to meet strict data‑sovereignty and security standards.
Ollama Web Search and Web Fetch API Release
Ollama has released a new web search and web fetch API to reduce hallucinations and provide LLMs with real-time web data via REST and library integrations.
OpenAI, Oracle, and SoftBank Expand Stargate AI Infrastructure
OpenAI, Oracle, and SoftBank have announced five new AI data center sites in the U.S., bringing Stargate's planned capacity to nearly 7 gigawatts and over $400 billion in investment.
Smol2Operator release: turning a small VLM into an open‑source agentic GUI coder
Hugging Face released Smol2Operator, a post‑training recipe that converts the 2.2 B‑parameter SmolVLM2‑2.2B‑Instruct model into an open‑source agentic GUI coder, with all code, datasets, and the resulting model publicly available.
Ollama New Model Scheduling Update
Ollama has introduced a new model scheduling system that replaces memory estimation with exact memory measurement to reduce crashes and increase GPU utilization.
Qwen3Guard Release: Real-time Safety Guardrails for LLMs
Qwen introduces Qwen3Guard, a family of safety guardrail models available in Gen and Stream variants to provide real-time, multilingual safety classification for prompts and responses.
CNA AI Newsroom Transformation
Channel NewsAsia (CNA) is integrating AI as a foundational backbone technology to enhance public service journalism through custom GPTs, disinformation analysis, and operational efficiency.
SchoolAI Platform Integration with OpenAI
SchoolAI has deployed an AI platform powered by OpenAI models to provide personalized student support and real-time teacher insights, reaching 1 million classrooms across 80+ countries.
OpenAI and NVIDIA Strategic Partnership for 10 Gigawatts of AI Infrastructure
OpenAI and NVIDIA have entered a strategic partnership to deploy at least 10 gigawatts of NVIDIA systems and an investment of up to $100 billion from NVIDIA to build next-generation AI infrastructure.
SyGra: A Low-Code Framework for LLM and SLM Data Generation
SyGra is a low-code/no-code framework developed by ServiceNow AI to simplify the creation, transformation, and alignment of datasets for Large and Small Language Models.
Hugging Face Gaia2 and Meta Agents Research Environments (ARE) Release
Hugging Face has released Gaia2, a read-and-write agentic benchmark, and the Meta Agents Research Environments (ARE) framework to evaluate and debug AI agents in complex, real-world simulated conditions.
DeepSeek-V3.1-Terminus Release Notes
DeepSeek has released DeepSeek-V3.1-Terminus, an update to V3.1 that improves language consistency and enhances the performance of Code and Search Agents.
Scaleway added as Hugging Face Inference Provider – capabilities, usage, and billing
Scaleway is now an official Inference Provider on the Hugging Face Hub, offering low‑latency, European‑hosted serverless access to frontier models with flexible billing options.
Ollama Cloud Models Preview
Ollama has introduced Cloud models in preview, enabling users to run large-scale models on datacenter-grade hardware while maintaining the same local toolset and API integration.
Grok 4 Fast release notes / what's new
xAI has released Grok 4 Fast, a cost-efficient reasoning model featuring a 2M token context window and a unified architecture for reasoning and non-reasoning modes.
Hugging Face RiskRubric.ai Announcement
Hugging Face has announced RiskRubric.ai, a standardized risk assessment platform that evaluates AI models across six pillars of transparency, reliability, security, privacy, safety, and reputation to provide comparable risk scores and letter grades.