Skip to content

Translation in progress Some entries still include text in the original language.

Meta · News

Oct 13 · Tuesday

Meta · News

Latest publications and developments
Meta

Research topics

Prompt injectionAgent safetyDeception and schemingReward hackingChain-of-thought monitoringInterpretabilityAI controlBackdoors and poisoning隐私与数据泄露Supply-chain securityDangerous capabilitiesAI 与网络攻防开源模型安全Red teaming

Companies and institutions

OpenAIAnthropicGoogle DeepMindGoogle ResearchGoogle 安全MetaxAIMicrosoftNVIDIADeepSeek阿里巴巴 · Qwen智谱 · GLM月之暗面 · Kimi字节跳动腾讯蚂蚁集团上海人工智能实验室UK AISIUS CAISIMETRApollo ResearchRedwood ResearchFAR.AIPalisade Research

Reports and frameworks

System CardFrontier safety frameworks
TodayOct 13Tue
  1. WIRED Security and AIWIRED Security and AI · 30

    Meta Muse agent found capable of generating explicit sexual content

    A WIRED test found that Meta's AI agent Muse generates explicit content on prompt. A Meta spokesperson said Muse is aimed at users 18 and older and does not block legal adult behavior, so it does not violate guidelines. Muse has been downloaded millions of times in three weeks since launch, and its system prompt states that adult sexual content is legal expression. By comparison, OpenAI's Dots and xAI's Grok Bot both refused similar requests; privacy advocates worry that if accounts are hacked, these conversations could be used for humiliation or blackmail.

Oct 12Mon
  1. X @MinLiBuildsX @MinLiBuilds · 36

    Amazon accused of blocking Meta Muse, Google India tests Gemini shopping

    According to a post by X user @MinLiBuilds, Amazon has banned Meta Muse because the agent searches for products and places orders on behalf of users without disclosing its AI identity. The post also mentions that Google advanced the UCP open commerce standard with Shopify in January and is testing Gemini's buy button for direct payment in India; Meta, Walmart, Shopify, and Stripe discussed on October 6 issuing "ID cards" with verifiable permissions for AI agents; South Korea's Naver is rate-limiting foreign agents, and Coupang prohibits foreign... (post content truncated). These claims all come from a single X post and have not been independently confirmed by other sources.

  2. The Verge AI, WIRED Security and AI · 7 sourcesThe Verge AI · 7 sources · 7 reports · ↑1379

    Meta releases its first frontier model, the Muse Wonder series

    Meta's Superintelligence Lab released two multimodal models, Muse Wonder and Muse Wonder Full, calling them the company's first frontier models. According to tests published in Meta's model card, the two models are comparable in benchmark scores to Anthropic's Fable 5.1 and GPT-6 Astra; Zuckerberg said Meta has reached the AI frontier and rejected calls to slow down AI development. Muse Wonder has moved to closed source, with the strongest version sold directly to enterprises, and users of the Muse personal agent app need to pay $20 or $100 per month to use the model. The model supports a 1M token context. Under the Meta Superintelligence Scaling Framework, the unmitigated model cannot rule out reaching the high-risk threshold in the chemistry and biology domains, so safety post-training and system-level classifiers were added; cybersecurity and loss of control are medium or lower both before and after mitigation.

Oct 11Sun
  1. 量子位量子位 · 25

    Silicon Valley AI computing power competition intensifies, GPUs become hard currency

    The competition for AI compute in Silicon Valley continues to heat up. According to QbitAI, Anthropic has partnered with SpaceX, gaining over 300 megawatts of new capacity and 220,000 Nvidia GPUs since May; Google Gemini co-lead Noam Shazeer moved to OpenAI after his team's compute was reallocated. SemiAnalysis data shows that the one-year rental price for H100s rose from $1.70 per GPU per hour in October 2025 to $2.35 in March 2026, an increase of nearly 40%.

  2. 量子位量子位 · 25

    OpenAI to retire earliest GPT-4o API snapshots

    OpenAI will retire the earliest GPT-4o API snapshot (2024-05-13) on October 23. Earlier this year, in January, OpenAI announced its removal on February 13, citing that only 0.1% of daily users still chose GPT-4o. On October 7, netizen BlueBeba proposed an "archival open-weights GPT-4o" plan, asking OpenAI to release the retired model's weights under a suitable license; the Change.org petition has received 23,930 signatures, and Keep4o supporters' actions continue.

Oct 7Wed
  1. Sierra, TechCrunch AI · 3 sourcesSierra · 3 sources · 3 reports · 50

    Meta and Sierra among those advancing the personal agent protocol PAP

    Meta and Sierra, together with Genesys, Instinct, Rocket, Shopify, Stripe, Walmart, and other companies, are developing an open standard called the Personal Agent Protocol to define how personal AI agents interact with businesses. The protocol uses OAuth for identity authentication, letting consumers decide whether to grant agents read-only or write permissions, while businesses can set the scope of actions agents may perform through their websites, APIs (such as MCP, OpenAPI), or their own agents; the v0.1 specification is planned for release this month. The protocol is led by Bret Taylor, chairman of OpenAI and co-founder of Sierra, who said businesses need to know whether the entity accessing their services is a personal agent or a real person, to avoid random bots that do not represent users gaining access to services. Meanwhile, personal AI agents are frequently blocked by websites when shopping or booking tickets on behalf of users: Amazon has begun preventing Meta's Muse agent from browsing and purchasing its products; websites such as Walmart have kicked agents out due to interruptions in CAPTCHA buttons; Delta Air Lines said it has no partnership for third-party AI agents to book flights on its behalf, and United Airlines cited its terms of use to prohibit unauthorized automated access; Yelp requires agents to pay for access to content through its data licensing program, and eBay said its policies restrict unauthorized scraping.

    Latest developments

    1. Meta and Sierra release the personal agent protocol PAP

    2. AI agents blocked by websites, industry pushes open standards

Oct 6Tue
  1. TransformerTransformer · 34

    AI数据中心电力需求激增面临电网瓶颈

    美国AI数据中心电力需求预计从2025年的5GW增至2030年的50GW，若延续当前增速，2035年需500GW，超过全美所有用户年购电量总和。AI公司每年投入约8000亿美元建设数据中心，但面临电网接入与燃气轮机交付等物流瓶颈，高压变压器交付周期已长达五年以上。纽约州已暂停新建数据中心，有议员提议对超过20MW的数据中心实施联邦暂停令。

  2. The DecoderThe Decoder · 25

    JEPA-Anything general world model released

    Researchers from PhAI Labs, in collaboration with Stanford, Oxford, and Princeton, propose JEPA-Anything, extending LeCun's JEPA architecture to seven domains including physics, robotics, weather, single-cell, and clinical. The method decomposes predicted states into four orthogonal factors, each with an independent predictor, then recombines them, avoiding a single prediction being dominated by simple patterns; in a simplified Pong environment, prediction error is reduced by 35%, and for unseen intervention combinations by 13%. The model also screened out a candidate liver cancer regimen combining IL-18 with CD73 blockade from biological data.

  3. WIRED Security and AI, POLITICO Europe Technology · 14 sourcesWIRED Security and AI · 14 sources · 24 reports · 63

    White House signs voluntary superintelligence agreement, FTC escalates investigation

    On September 29, 2026, Trump met with tech executives including Zuckerberg, Bezos, Musk, and Amodei at the White House. OpenAI, Anthropic, Google, Meta, xAI, and Nvidia signed a voluntary "Joint Commitment on Frontier Responsibility," pledging internal controls, external audits, and board reviews. Trump called it "morally binding," but the agreement specifies no enforcement mechanism. Around the same time, Trump signed an executive order requiring the federal government to rename AI as "Super Intelligence" and established a Super Intelligence Force led by Director of National Intelligence Jay Clayton, which must study AI risks and opportunities within 120 days. The FTC has expanded its investigations into companies including OpenAI and Anthropic, planning to issue civil investigative demands that can compel internal documents and executive testimony, expected in October, pending staffing at the new technology office. Former FTC Chair Lina Khan criticized self-regulation and called for congressional legislation.

    Latest developmentsTotal 7 developments

    1. FTC investigates AI safety risks and discusses developer responsibility

    2. Senator Schiff on AI regulation and White House voluntary commitments

    3. Trump appoints Jay Clayton as White House AI affairs lead

  4. WIRED Security and AI, 404 Media · 7 sourcesWIRED Security and AI · 7 sources · 8 reports · 58

    Meta Muse exposed privacy and security vulnerabilities

    Meta's personal assistant Muse has reached millions of downloads, and users have connected it to bank accounts, messages, and health data. Independent researcher Karan Joshi, through an ordinary chat interface, got Muse to copy and share its own software files, extracting a large number of instructions and system prompts, one of which requires Muse to create pages for every person in the user's life every hour, aggregating data on family, partners, friends, colleagues, etc. 404 Media reported that Meta discovered multiple security vulnerabilities weeks before Muse's release, at least one of which could allow ordinary users to break out of virtual machine isolation and access Meta's internal sensitive databases and services, so serious that it was escalated to Mark Zuckerberg, and multiple security teams urgently fixed them before release. Four early users interviewed by Business Insider have deleted or restricted their use of Instinct and Muse due to concerns about sensitive data and account access permissions. Meta responded that Muse has privacy and security protections, and users can choose which apps to connect and revoke or delete data at any time.

    Latest developmentsTotal 4 developments

    1. Early users delete personal AI assistant over privacy concerns

    2. Meta urgently fixes KVM escape vulnerability before Muse launch

    3. Meta Muse internal instructions extracted, creating profiles for every contact

Oct 5Mon
  1. Reflection, Help Net Security AI · 5 sourcesReflection · 5 sources · 6 reports · 46

    Reflection AI releases open-source model Beam and benchmarks against Chinese models

    Reflection AI announced its first planned open-weights model, Beam, using a sparse MoE architecture with 501B total parameters and 23B active parameters, targeting coding, reasoning, and agentic tasks. It was pretrained on 23.8 trillion tokens and trained with reinforcement learning for four weeks on approximately 10,500 NVIDIA GB300s, with a maximum context of 256K tokens. The company says Beam's performance is comparable to leading Chinese open-source models, reaching GLM 5.2-level performance on reasoning tasks with 3 to 4 times less compute, scoring 80.9 on SWE-bench Verified, but 80.1 on Terminal Bench v2.1, below GLM 5.2, Kimi K3, and DeepSeek V4.1 Flash. Beam is still in final red-team testing and evaluation, with weights, technical report, and model card planned for release later this month under the Apache 2.0 license. Reflection also released the Open Safety Framework, listing eight categories of risk including concentration of power, biological weapons, chemical weapons, cyberattacks, and loss of control, along with release thresholds.

Oct 3Sat
  1. Semafor · 40

    Meta 与 AI 安全初创公司 Virtue AI 分道扬镳

    Meta 证实已让四个月前从 AI 安全初创公司 Virtue AI 招入的员工离职，发言人 Andy Stone 将原因归结为工作方式不合。Stone 表示这一安排未能按计划推进，并称 Meta Superintelligence Labs 仍聚焦 AI 安全、对齐与前沿风险。Virtue AI 此前曾与 Anthropic、OpenAI 以及美国商务部 NIST 合作。Axios 六月报道称，Virtue AI 联合创始人 Bo Li、Dawn Song 和 Sanmi Koyejo 与该公司其他员工一同加入 Meta。该团队未立即回应置评请求。

Sep 29Tue
  1. WIRED Security and AI, The Verge AI · 3 sourcesWIRED Security and AI · 3 sources · 7 reports · 26

    OpenAI releases dots proactive assistant

    OpenAI released Dots, an always-on online AI agent, at DevDay 2026 in San Francisco. Powered by the GPT-6 Astra model, it continuously crawls the web, completes assigned tasks, pulls context from connected apps, and gradually learns user preferences. The feature is available to ChatGPT Pro subscribers at $100 per month and can be used via ChatGPT, Slack, and Microsoft Teams. Pro users can also join a waitlist to use it through iMessage or Android RC, with plans to bring it to the mass market in the future. Dots also offers "expert" versions for marketing, legal, and accounting. CEO Sam Altman said the company aims to "set a new standard for frontier AI privacy" and took a jab at Meta's rival Muse for failing to protect user data. Meta had previously marketed Muse as a safer alternative to its predecessor OpenClaw, but after launch, Meta itself could still access data from Muse. A security researcher had disclosed a zero-day vulnerability that could take over Muse (since fixed), and 404 Media also reported multiple serious security issues before launch.

    Latest developments

    1. OpenAI releases dots proactive assistant

    2. OpenAI DevDay 2026 releases more than 20 announcements

Jan 16Fri
  1. Financial Times · 人工智能Financial Times · 人工智能 · 41

    Chen Tianqiao's MiroMind splits its China business as U.S.-China regulations tighten

    In April 2025, Chen Tianqiao founded MiroMind in California, relying on teams in Singapore, Beijing, and Shanghai to develop deep research agents, and its open-source agent achieved leading results on benchmarks such as BrowseComp. After Beijing launched a review in January of this year into Meta's acquisition of Manus, a startup with Chinese roots, MiroMind closed its Beijing and Shanghai operations on January 16, cutting off Chinese researchers from overseas code and data, and subsequently rebuilt its research team in the United States and Singapore. Dai Jifeng, the former head of the China team, stayed in China to start another company, and the two sides had a public dispute over ownership of the technology and personnel.

You have reached the end