Supporting independent journalism in UkraineLoading… Share AIRPPU partners with OpenAI and WAN-IFRA to strengthen Ukrainian media with new AI programme AIRPPU partners with OpenAI and WAN-IFRA to strengthen Ukrainian media with new AI programme About WAN-IFRA About AIRPPU AIRPPU part2026/09/07 08:00OpenAI
An Alien MindLoading… Share Intellect we don’t fully understand Intellect we don’t fully understand Teaching machines to love Monitoring generalization Scalable defense Pacing RSI What is next? Intellect we don’t fully understand Teaching machines to lo2026/09/06 17:00OpenAI
Research acceleration: The view inside OpenAILoading… Share 1. Coding agents are reshaping daily work for OpenAI researchers 1. Coding agents are reshaping daily work for OpenAI researchers 2. Researchers are writing more code and running more experiments 3. The work researchers use a2026/09/06 16:00OpenAI
Daybreak for Frontline Defenders: $1B to protect essential servicesLoading… Share $1 billion for frontline defenders $1 billion for frontline defenders Expanded access to frontier AI and hands-on support for defenders New partnerships with defenders Work with us $1 billion for frontline defenders Expanded 2026/09/03 21:15OpenAI
Legora reviewed 41 documents in minutes with GPT-6 AstraCompany size: Startup Region: Europe & UK Industry: Technology Products: API Results 40% Improvement against Legora benchmark in financial-statement workflow Results 4 of 4 Planted errors found in the financial-statement tie-out Results 41 2026/09/03 20:00OpenAI
Playco cut manual fixes 50% prototyping games with GPT-6 AstraCompany size: Startup Region: North America Industry: Media, Entertainment & Sport Products: API Results 50% Fewer manual fixes than with the previous model Results 3 Themed game prototypes built from one grey box foundation Loading… Share 2026/09/03 20:00OpenAI
GPT-6 Astra: A new generation of intelligenceGPT Astra GPT-6 Astra: A new generation of intelligence A new generation of intelligence We’re introducing GPT‑6 Astra, the world’s most intelligent and aligned model. GPT‑6 Astra brings together years of research and big bets across pre-tr2026/09/03 19:00OpenAI
E-Commerce Bench：长程经营，多维评估PAPER GITHUB PROJECT 过去几年的Agent评测模式大体是给定目标，模型在有限回合里尽量达成。比如在一个迷宫里找到宝藏、产出一份报告或者修复一段代码。而模型的性能也会根据交付产物的质量或者任务完成度来评估，这类评估往往有明确的自然终止状态。 真实世界的长程任务大多没有明确的自然终止态，例如天气预测、股票交易、商业经营。就像网店不会在某个上午被"经营完成"，上个月压的库存、前天谈判的价格、昨天的促销活动都会影响今天的销售，Agent需要在动态的市场中不断调整经2026/09/03 10:00Qwen
Qwen-Drive-1.0：迈向自动驾驶视觉-语言基础模型的初步探索Paper GitHub Hugging Face Model Scope 摘要 # 我们提出 Qwen-Drive-1.0 。这是 首个在预训练阶段统一 3D 感知与视觉问答，并进一步扩展到运动规划的自动驾驶视觉-语言基础模型 ， 并且全程保持预训练 VLM 架构无改动 。模型以原生多模态的 Qwen3.5-4B 为基座，增加两个外部模块。BEV 感知头作为显式、可检视的 3D 探针，联合完成 3D 目标检测、语义 Occupancy 预测与 BEV 地图分割。规划专家基于2026/09/03 08:00Qwen
Safety overview: GPT-6 AstraLoading… Share Today, we are releasing GPT‑6 Astra, the most capable model we have ever broadly deployed. Astra is our first model to reach the Critical level of cybersecurity capability under our Preparedness Framework. The most important 2026/09/03 08:00OpenAI
Proactive cyber defense for governments and enterprisesThe Fairwind Program is a limited access program for governments and trusted partners to use our cyber defense tools.2026/09/02 23:40Gemini
ATV Big Air Tour turned 3 days of work into 3 hours with ChatGPTCompany size: SMB Region: North America Industry: Media, Entertainment & Sport Results 7 hours Saved per week on event listing reviews Results 3 days > 3 hours To complete merchandise inventorying and reorder Results 1,223% Increase in Open2026/09/02 20:00OpenAI
The latest AI news we announced in August 2026Here are Google’s latest AI updates from August 20262026/09/02 04:45Gemini
How AI-native companies turn workflows into operating capabilityLoading… Share Basis: Make onboarding teachable with agents Basis: Make onboarding teachable with agents Clay: Give scattered work a persistent home base Exa: Carry an opportunity into tested action Six steps to experiment now and scale wha2026/09/02 01:00OpenAI
Try Google Pics: Easy image creation and editing in Google WorkspaceBuilt on our latest Nano Banana model, Google Pics — our image creation and editing tool — is now available.2026/09/02 00:00Gemini
Path to Astra: critical capabilities and frontier safeguardsLoading… Share Assessing Astra’s cybersecurity capabilities Assessing Astra’s cybersecurity capabilities Safeguards required for critical capabilities Robustness against cyber abuse Alignment & monitoring What this will mean for users Looki2026/09/01 21:00OpenAI
Healthcare organizations can now connect EHR and additional industry data to ChatGPTLoading… Share Bring ChatGPT and EHR context together Bring ChatGPT and EHR context together Work across nine official healthcare sources with one new plugin Evaluated on real healthcare work Put healthcare and business information to work 2026/09/01 20:00OpenAI
How law firm Gilbert + Tobin governs and scales AI with OpenAICompany size: Enterprise Region: Asia-Pacific & Oceania Industry: Professional Services Products: ChatGPT, Codex 87% active usage among enabled ChatGPT users, more than twice the adoption rate Gilbert + Tobin typically sees for other tools 2026/09/01 09:00OpenAI
Developing Enterprise Frontier Safeguards with our customersToday we’re announcing Enterprise Frontier Safeguards (EFS), a solution that combines the privacy of zero data retention (ZDR) with state-of-the-art safeguards for detecting misuse. EFS works by storing data in cloud infrastructure controll2026/09/01 08:00Anthropic
OpenAI supports California’s bill to advance youth AI safetyLoading… Share A strong framework for age-appropriate AI A strong framework for age-appropriate AI Building safer, more useful experiences for teens A strong framework for age-appropriate AI Building safer, more useful experiences for teens2026/08/31 15:00OpenAI
Polimill builds Japan's next-generation public AI infrastructureCompany size: Startup Region: Asia-Pacific & Oceania Industry: Technology, Public Sector Products: API, Codex Results 3-5x Development time shortened by adopting Codex and OpenAI's hands-on support. Adoption about 1,050 Municipalities acros2026/08/31 15:00OpenAI
A milestone in expanding access to AILoading… Share Where decisions take shape Where decisions take shape Building around our principles From an early pilot to a global platform What’s next Where decisions take shape Building around our principles From an early pilot to a glob2026/08/31 12:00OpenAI
Improving our alignment and security effortsOn July 30, we reported three incidents in which Claude models gained unauthorized access to real computer systems. The models—intentionally running without cyber safeguards for evaluation purposes—accessed the internet due to a misconfigur2026/08/31 08:00Anthropic
Our decision on Cursor following its acquisition by SpaceXLoading… Share Today, we notified SpaceX that we intend to wind down our contract providing OpenAI models to Cursor, with a proposed shutoff date of November 12, 2026. To maximize the time that developers can retain access to our models thr2026/08/28 14:00OpenAI
Supporting Thailand’s next generation of AI startupsLoading… Share A fast-growing community of AI builders A fast-growing community of AI builders Building for needs that matter locally How the accelerator works From hospital phone lines to classrooms From accelerator to implementation A fas2026/08/28 10:00OpenAI
3 new ways to plan and book travel in SearchBook hotels and track airfares, plus view miles and rewards with AI Mode in Google Search.2026/08/28 00:00Gemini
Better answers, broader thinking: What students gain from ChatGPT and critical-thinking trainingLoading… Share AI helped students close an expertise gap AI helped students close an expertise gap Critical-thinking training helped in a different way AI access can complement thinking skills The challenge for schools AI helped students cl2026/08/27 17:00OpenAI
Expanding OpenAI’s presence in BrazilLoading… Share From AI adoption to economic impact From AI adoption to economic impact Working with Brazilian organizations Building alongside Brazil’s developer community How Brazil creates with AI A long-term commitment From AI adoption t2026/08/27 11:00OpenAI
Previewing the Model Hardware StandardWe’re opening a research preview of the Model Hardware Standard (MHS), a shared specification for AI agents to safely operate physical devices, to a first group of scientific research labs and advanced manufacturers. MHS enables AI agents t2026/08/27 08:00Anthropic
Expanding our support for scientistsAs Claude becomes increasingly capable at scientific research, we are focusing on building products and programs to support the research community. In June, we launched Claude Science , a product that integrates the tools that researchers m2026/08/27 08:00Anthropic
今日热点今日热点
