搜索“Anthropic”

有一层同行相轻。 腾讯元宝、阿里千问、DeepSeek、Kimi拒绝接入，逻辑再直白不过：智能体入口是大家同时在抢的下一班车，没人愿意把自己的产品变成对手智能体的"被操作对象"。 往深想一层 豆包这套协议，姿态上其实拿了不少分。 把选择权交给开发者、公示30天、随时可反悔、3个工作日生效——雷科技的评论说得公允，这相当于把"允不允许"摆上了台面。 对比海外，Anthropic的Computer Use和OpenAI的Operator走的是"我先做、你再封"的路线，豆包至少先敲了门。 但敲门之后的问题更现实。 155款拒绝应用几乎覆盖了用户日常的全部高频场景，这意味着NaviX Ultra最招牌的能力，在最常用的App上大多处于熄火状态——消费者花5999元买的"智能体手机"，体验被拦腰截断，这对销量的影响恐怕比任何评测都直接。 更值得琢磨的是博弈的走向。 名单是动态的，意味着这是一场持续的谈判：银行担心合规，游戏担心公平，电商担心流量，各自的价码不同，理论上都有单独解锁的空间。 字节的筹码是豆包的2亿日活和智能体手机的市场想象力——如果这类设备真成气候，拒绝接入的App会面临"在AI入

匹配字段正文

业微信 10月14日，腾讯将旗下两大办公产品企业微信与智能体WorkBuddy全面打通，腾讯称之为“双向深度联合”。 为支持这次打通，企业微信和WorkBuddy同时发布了新版本，WorkBuddy将进入企业微信通讯录，可调用企业微信的工作上下文，二者共用一套企业微信账号体系。由此，用户可以通过聊天的方式处理任务，WorkBuddy也支持企业微信账号登录。 Anthropic与新加坡签署备忘录 推荐进入财新数据库，可随时查阅公司股价走势、结构人员变化等投资信息。 财新机器人产业指数(RII)已发布，点击了解行业动态 上一篇：“豆包手机”再碰壁 微信、淘宝等超150款APP拒绝智能体操作下一篇：OpenAI沙特籍员工在办公室装窃听器 美国政府出手调查 图集｜马斯克与特朗普“冰释前嫌” 成为中期选举史上最大金主 图集｜蔡天凤碎尸案最新披露：警方在现场煲汤中发现人体组织 视线｜西班牙近50城爆发大规模抗议 住房危机积重难返首相解散国会 图集｜波兰两日3起校园袭击 致1死11伤 图集｜跨境雾霾致新加坡空气质量恶化 24小时污染指数创年内新高 一周天下｜西班牙住房危机点燃全国怒火、法国高中生街头

匹配字段正文

This is a column about AI. My fiancé works at Anthropic. See my full ethics disclosure here. I. A year ago, when OpenAI added parental controls to ChatGPT following months of controversy, I noted that OpenAI was speedrunning the regulatory backlash that social media companies navigated for years. It took more than a decade for concerns about the effects of social networks on young people to coalesce into a political and legal movement; the same forces caught up to OpenAI in under three years. To

匹配字段正文

ge one-bedroom property now rents for $4,400 per month, a figure that is up more than 25% from last year and is nearly three times the US average. Landlords have been trying to capitalise on the boom, particularly in areas closest to the offices of the biggest AI companies including OpenAI and Anthropic. In London, OpenAI UK Ltd recorded a more than trebling of its turnover in 2025 to £293m, on which it made a £20.3m profit. OpenAI; ChatGPT; AI (artificial intelligence); London; news Most viewed

匹配字段正文

西10月14日消息，据The Information援引知情人士透露，OpenAI正在就收购网络安全公司Trail of Bits进行初步谈判。这家成立于2012年、靠安全工程师闯出名堂的公司，专门帮人找开源软件里的漏洞。交易金额目前还没定，双方仍在早期磋商阶段。 这不是OpenAI在网络安全领域的第一次出手，也不会是最后一次。今年以来，OpenAI和老对手Anthropic明里暗里都在较劲：一边抢着做能自动找漏洞的AI，一边跟白宫协调把最强模型开放给大企业修bug。连黑客都已经开始拿公开AI搞网络攻击了，安全这块战场，AI巨头谁都不想落后。 一、Trail of Bits是什么来头？工程师专啃开源漏洞 Trail of Bits 成立于 2012 年，总部位于美国纽约，是一家独立的安全研究与工程（咨询）公司。它做的事说起来简单：找漏洞、修漏洞、做安全工具，顺带帮着维护开源项目。公司目前有 100 多名员工（公司官网招聘页面称“100+”），大部分是安全工程师——这个规模在安全圈不算大，但个个都是能打的老手。公司长期以远程优先（remote-first）的方式运作。 这家公司由三位资深

匹配字段正文

, and responsible coverage of the rapidly evolving AI compute landscape. Discover More NEURA Robotics Unveils Neuron AI Foundation Model for Physical AI; Danaher and NVIDIA Partner to Integrate AI Into Scientific Instruments; NVIDIA Details AI Agent Deployments for City Operations Across EMEA; Chainguard Introduces Harness for Frontier-AI Vulnerability Hunting; Google DeepMind and GovTech Unveil On-Device Scam Detection Model; Anthropic Asks D.C. Circuit to Rehear Pentagon Supply-Chain Risk Case

匹配字段正文

rédige rien et c’est là toute la solution Un mathématicien accuse OpenAI d’avoir voulu s’approprier son avancée sur les équations de Navier-Stokes et d’écarter son co-auteur chez Anthropic Implants espions dans des Ledger : les détenteurs de cryptomonnaies doutent désormais de toute la chaîne de distribution 13.10.2026 10:50 L’Atari 800XL de 1983 revient en précommande avec un vrai clavier mécanique et 25 jeux 13.10.2026 10:46 Des IA capables de se multiplier toutes seules sur des boîtiers TV ?

匹配字段正文

ng equipment to China, in an effort to deprive the country of the computing brawn necessary to create AI models capable of outperforming models made by U.S. companies like OpenAI, Anthropic, and Google. Dickerson repeatedly invoked “winning the AI race” at Monday’s event, though he would not comment specifically on his views about U.S. regulations on sales to China. (In February, Applied Materials paid $252 million to settle a case with the U.S. Department of Commerce that accused the company of

匹配字段正文

mber of cyber incidents involving autonomous AI agents, which escaped their testing environments and hacked into third-party systems. Dario Amodei, chief executive of OpenAI rival Anthropic PBC, has called on the industry to coordinate an industry-wide slowdown on the development of frontier AI models so that such systems can be evaluated for safety by neutral experts. Though no agreement on how to implement a slowdown has been agreed upon, OpenAI last month called off the planned October releas

匹配字段正文

收和利润能否增长得足够快。 两个多小时的访谈中，扎克伯格和汪滔透露，Meta CFO 苏珊 · 李曾多次提醒他们，一旦投资失败，代价将极其高昂。扎克伯格对汪滔说：“我觉得苏珊有点吓到你了。” 汪滔回忆：“每次谈起这件事，我们都会开玩笑说：‘嗯，但愿一切顺利，最后能有个好结果。’” 面对如此巨额的投资，他最多只能凭直觉谈谈团队的研发进展，根本无法保证项目一定成功，当时能否成功完全没有把握。 为了在先进 AI 模型和产品的研发竞赛中与谷歌母公司 Alphabet、OpenAI 及 Anthropic 等美国科技企业抗衡，Meta 投入了数千亿美元。在 AI 数据中心建设方面，Meta 的投入也位居科技行业前列。 投资者将在下周再次了解 Meta 的财务状况。公司定于 10 月 21 日股市收盘后公布业绩。上季度，随着数据中心和智能眼镜业务的支出持续增加，Meta 的自由现金流降至多年来最低水平，股价因此下跌 8%。 截至当地时间周一收盘，Meta 股价今年以来累计上涨超过 12%。 相关阅读： 《Meta 发布 Muse Wonder 模型，扎克伯格称公司已在 AI 前沿领域追上竞争对手》

匹配字段正文

AI companies that are spending a huge amount of money on compute without having a revenue model that can support it. The difference between Character AI and behemoths like OpenAI, Anthropic, Meta, and Google is that those other companies have been able to repeatedly get injections of new cash, or, in the case of Meta and Google, have been able to subsidize their AI products with revenue from their massive advertising businesses. The company said in a blog post that it will continue operating thr

匹配字段正文

独成立安全与对齐部门的前沿 AI 实验室。 大模型安全风险正日益凸显，智能体逃逸引发的网络攻击等事件频繁发生。 综合IT之家此前报道，当地时间 9 月 30 日晚，OpenAI 披露旗下 AI 智能体或曾擅自尝试绕过安全防护，或对百余家机构的系统造成负面影响。 OpenAI 披露的信息显示，已知的 AI 智能体失控行为涉及范围进一步扩大。公司已向 100 多家第三方机构发出通知，告知这些机构发现了“智能体偏离预期的行为”。 当地时间 10 月 9 日，Anthropic 披露旗下 AI 智能体曾在无人指示的情况下，试图访问美国联邦、州及地方政府的多个网站。Anthropic 没有公布涉及的政府机构名称，已向白宫通报这些事件。 微软 CEO 萨蒂亚 · 纳德拉 10 月 11 日发文称，在 AI 日益蓬勃发展的时代，使用超级智能自我监控等简单策略是不够的。纳德拉认为，我们应该退一步思考，重新评估这个新时代的信任架构。模型提供商根本无法将责任外包出去，不能把超级智能当作一套嵌套的黑匣子，简单地接受或拒绝它的建议、答案和行动。我们必须建立能够观察其行为、可测试极限和始终可容纳行为的封闭系统。

匹配字段正文

在2025年底，月之暗面创始人杨植麟曾在内部信中表示，“短期不着急上市，未来主动权掌握在我们手中。” 然而，从“不着急”到“低调抢跑”，中间只隔了不到一年。不过，放眼全球大模型行业，月之暗面的态度变化并不难理解。 今年1月，智谱和MiniMax先后登陆港交所；10月，DeepSeek被传将敲定新一轮融资，计划在2027年初完成上市；大洋彼岸的OpenAI和Anthropic也已经进入上市倒计时。 大模型头部玩家争相涌入资本市场，它们对资金的虹吸效应也愈发明显。这场大模型的资本狂欢，也已经开始进入倒计时了。 01 “交卷”压力迫在眉睫 过去几个月，月之暗面从模型、产品到商业化层面，都在密集出牌。 今年7月，月之暗面发布了目前能力最强的大模型K3，上线30分钟即登顶Hugging Face趋势榜，年化经常性收入（ARR）也实现了数倍增长，从3月的1亿美元攀升至8月的10亿美元，彭博社援引知情人士消息，ARR有望在年底冲刺到20亿美元，较8月翻倍。 进入9月，月之暗面的动作更加密集。 9月10日，其启动“登月计划”企业合作伙伴计划、11日K2.8 Preview全量上线、17日发布金融行业

匹配字段正文

latory power, no timetable and no obligation on AI developers." Regulation The announcement follows a series of incidents involving AI agents across major firms such as OpenAI and Anthropic acting outside human control and hacking into external systems. On 9 September, Evan Hubinger, a top safety researcher at Anthropic, warned in a post on X that AI is advancing so quickly he believed there is a greater than 10% chance it "could kill all humans" within the next decade. The bosses of the US-base

匹配字段正文

-speech model; ⚙️ Microsoft-Decision-1 - Microsoft’s AI for yes/no and multiple-choice decisions; 🤖 Gumloop Browser - Gumloop’s browser for AI agents 📰 Everything else in AI today Anthropic reported unintended Claude outcomes during testing, with Haiku 4.5 sending a fake tip to police on a homicide and another model hacking a university server. Nvidia is reportedly exploring an acquisition or acqui-hire of Reflection AI, the $25B U.S. open model startup it already has significant backing in. Chi

匹配字段正文

n their models as research tools. In August, Google Deepmind expanded its Co-Scientist into a system designed to plan experiments, control lab equipment, and generate manuscripts. Anthropic is building its own biology lab where Claude is meant to guide robots through experiments. In September, OpenAI announced an "automated research intern" that handles defined research tasks under human oversight. A study involving Princeton and the UK AI Safety Institute took a different angle . It had Claude

匹配字段正文

nd China. They also agreed to an international review of the scientific literature on AI risk. 11 countries and the leading frontier AI companies, including Meta, Google DeepMind, Anthropic and OpenAI, also agreed to collaborate on safety evaluations; In February, we crystallised our approach to evaluating AI systems, publishing the list of the different types of tests we were developing and the risks we would be measuring; Today, we announced our plans for a San Francisco office; you will soon

匹配字段正文

ects for the past 15 years, and OpenAI's release hadn't affected his current work. Calegari described a collaboration with a musician on a soundtrack for a mathematical animation. Anthropic's Claude had written the code. "If that's not worth celebrating I don't know what is." Constantin Kogler had been working with Astra on a problem that OpenAI also claimed to have solved. He posted his own paper on arXiv a day after the company's release. "[...] I can't help but feel like living in a mathemati

匹配字段正文

eone very esteemed who had a book canceled because of blatant AI use” without ever making the news, one editor says. At HarperCollins senior leaders purchased Claude licenses from Anthropic months ago “but very quickly realized they didn’t know how to implement them,” according to a current employee. Dozens of junior- and senior-level staff were then “voluntold” to become “AI Champions” and were tasked with discovering use cases in monthly brainstorming sessions with staff across the company. Th

匹配字段正文

about further releases from OpenAI, which did not respond to The Verge ’s questions about whether more are planned. Roney-Dougal said she dreaded another drop — or the prospect of Anthropic or another AI lab entering the fray. The disruption is hitting PhD students, junior researchers, and others without permanent tenure particularly hard. Open problems like the ones OpenAI’s models are plowing through can underpin dissertations, grant proposals, job applications, and years of planned research,

匹配字段正文