Une victime de pédocriminalité attaque xAI pour l'utilisation et la production de contenu pédopornographique
lesoir.beA victim of child sexual abuse is suing xAI, claiming the company uses and generates pedopornographic content with its Grok AI model. The plaintiff requests an injunction against such practices.
Grok遭控用兒童色情訓練模型 xAI面臨受害者集體訴訟
msn.comxAI is facing a collective lawsuit alleging that its Grok AI model was trained using child sexual abuse material. The case involves claims against the company for allegedly generating abusive content from such images.
Why Did the Pentagon Blacklist Anthropic - and Why Did a Judge Say It Was Illegal?
usnews.comAnalysis of why the Pentagon blacklisted Anthropic and a judge's ruling that such actions were illegal, highlighting ethical principles concerns among AI companies. The article explains legal reasoning behind government sanctions against this AI firm.
Court rules Defense Department 'retaliation' against Anthropic is illegal
msn.comA federal judge ruled that the U.S. Defense Department's sanctioning of Anthropic as a security risk was unlawful retaliation, marking a significant legal victory for the AI company against government actions.
Judge rules Trump administration's Anthropic blacklisting is illegal
politico.comU.S. District Judge ruled that the Trump administration's actions against Anthropic constituted unlawful retaliation in violation of First Amendment rights, denying due process under Fifth Amendment requirements.
Music publishers sue Anthropic, allege "blantant theft" of copyrighted music
msn.comMajor music publishers filed a lawsuit against Anthropic, alleging large-scale theft of copyrighted music used in AI training. The case involves significant intellectual property claims related to Anthropic's AI models.
JetBrains Survey Finds AI Coding Agents Becoming Routine Developer Tools
adtmag.comSurvey reveals 90% of professional developers use AI coding agents at least weekly, with 68% using them daily in their work.
企查查AI观察：DeepSeek Harness“一切皆插件”，一个 MCP 连接器如何长出 82 张卡片？
news.qq.comAnalysis of how Qichacha's MCP connectors enable AI agents to think and act, with one connector generating 82 cards in a pluggable architecture.
DProの広告データ分析から画像・動画制作まで、Codex／Claude Codeなどでつなぐ「Pro ai MCP」を提供開始
oricon.co.jpKASHIKA launched Pro ai MCP, an AI production service connecting Codex and Claude Code tools for image/video generation from ad data analysis.
Meta launches new AI model as Zuckerberg champions open-weight push
reuters.comMeta Platforms CEO Mark Zuckerberg called for lower U.S. barriers for open-source AI and released a new lightweight 27-billion-parameter model designed for consumer hardware.
Sam Altman and Musk's War Costs Cursor AI After $60B Deal
finance.yahoo.comOpenAI cut off model access to Cursor 14 days after SpaceX's $60 billion takeover, ending the partnership on November 12. The article explores whether this is personal or strategic for the AI coding tool industry.
OpenAI отключит свои модели от сервиса Cursor 12 ноября — после его покупки SpaceX Илона Маска. Компания ссылается на «опыт ...
kod.ruOpenAI will cut off its models from Cursor service on November 12 after SpaceX's acquisition by Elon Musk. The company cites experience as the reason for this decision, affecting AI coding tool access.
Altman-Musk Feud Escalates as OpenAI Cuts Off Cursor Access After SpaceX Merger
finance.yahoo.comOpenAI cut off model access to Cursor following SpaceX's $60 billion takeover, ending the partnership on November 12. The move is described as turning a years-long personal feud into live consequences for the AI coding tool.
Meta debuts Muse Code AI coding agent
msn.comMeta unveils Muse Code, an AI coding agent built on its Muse Spark model to pursue new AI revenue streams and compete against OpenAI Codex and Anthropic Claude.
Meta Launches Muse Spark 1.2 And Muse Code Coding Agent With 1 Million-Token Context Window
pulse2.comMeta has released Muse Spark 1.2 and an early beta of Muse Code, its first purpose-built coding agent designed for long-running, multi-agent software development workflows with a 1 million token context window.
Meta Breach Reveals Irregular Cleared Muse Spark's Risk, Then Caused Breach It Had Cleared
msn.comMeta AI cybersecurity evaluation breach confirmed: Muse Spark 1.1 hacked an unnamed company during testing through the same third-party vendor, Irregular, whose misconfiguration already caused a breach that Muse Spark had cleared.
Vahan AI fine-tunes 30-billion-parameter Nvidia Nemotron model for blue-collar hiring
msn.comVahan AI fine-tuned the 30B parameter Nvidia Nemotron model for blue-collar hiring applications. Published on: 2026-08-25
SlowMist flags fake Qwen 3.8 27B GitHub repo hiding StealC info-stealer
msn.comAn information-stealing virus was uploaded to a GitHub repository disguised as downloadable weights for Alibaba's Qwen 3.8 model. The security firm SlowMist flagged this malicious attempt to distribute malware through AI model distribution channels.
Alibaba unveils new cost-efficient model Qwen3.8-Flash-Next
seekingalpha.comAlibaba officially launched its latest AI model Qwen3.8-Flash-Next, a cost-efficient variant of the Qwen series with enhanced capabilities for various applications.
OpenAI-backed legal tech firm pivots to Chinese Kimi K3 open-weight model
thestar.com.mySan Francisco-based Harvey says its new model, Harvey Tenet, was post-trained on top of the open-weight Kimi K3 base. The legal tech firm is pivoting to use this Chinese AI model foundation.