法官在这起案件中裁定，Anthropic 使用受版权保护的内容训练模型合法；但通过盗版方式获取内容本身是违法的。 最终这起诉讼导致 Anthropic 被罚款 15 亿美元（IT之家注：现汇率约合 104.29 亿元人民币），最终受影响的作家平均在每部作品分到了 3000 美元（现汇率约合 20857 元人民币）。 据报道，这些音乐公司在周三发布的声明中表示，索赔金额或超过 30 亿美元（现汇率约合 208.57 亿元人民币），将成为美国史上规模最大的 非集体版权诉讼案之一。 值得注意的是，此前 Anthropic 曾被一群小说作家起诉，对方指控这家 AI 公司使用受版权保护的作品训练 Claude 等产品。 IT之家 1 月 30 日消息，据科技媒体 TechCrunch 昨天报道，Concord Music Group、华纳音乐集团（Universal Music Group）等一批音乐出版商 已对 AI 公司 Anthropic 提起诉讼，称其非法下载超 2 万首受版权保护的乐曲，涵盖乐谱、歌词等。 ...

来源：环球网 【环球网科技综合报道】12月23日消息，据彭博社报道，一场版权诉讼在美国科技与文学领域掀起波澜。包括普利策奖得主 记者约翰·卡雷鲁在内的多位作家，正式向美国加州北区地方法院提起诉讼，将六家人工智能行业巨头推上了被告席，指控它 们使用作家们书籍的盗版副本来训练大型语言模型。 此次被起诉的六家公司分别为Anthropic、谷歌、OpenAI、Meta、xAI及Perplexity AI公司。诉状中，作家们言辞激烈地称这 些公司犯下了"蓄意盗窃行为"。值得一提的是，这是针对xAI训练过程提起的首起版权诉讼，同时也是作者对Perplexity发起 的第一起诉讼，在行业内具有标志性意义。 诉状详细披露了这些人工智能公司的侵权行径。据称，它们从LibGen、Z - Library和OceanofPDF等臭名昭著的盗版书籍"非法 影子图书馆"中，下载了原告作家们的著作盗版副本。这些公司的侵权行为并非仅止于此，在完成非法下载书籍这一初步侵 权步骤后，它们在训练大型语言模型或者对产品进行"优化"的过程中，又制造了更多的书籍副本，形成了二次侵权。作家们 痛心地表示，他们的心血之作"如今支撑着价值数十亿美元的产品生态 ...

Core Viewpoint - The article discusses the potential copyright crisis brought about by generative AI, highlighting how even a 6-year-old can create potentially infringing content in seconds using AI tools, which poses significant challenges for the intellectual property (IP) industry [1][12][14]. Group 1: AI and Copyright Challenges - A 6-year-old child was able to create a fully interactive website called "Bedtime Story Weaver" using Google AI Studio in under 2 minutes without any coding experience [6][5]. - The website allows users to generate complete stories and illustrations based on simple inputs, demonstrating how AI can democratize content creation [6][8]. - The child generated a story featuring characters from major franchises like Sonic and Mario, which raised immediate concerns about copyright infringement [10][12]. Group 2: Implications for IP Professionals - Jonathan Menkes, an IP lawyer, emphasized that the ease of generating infringing content with AI tools indicates that many in the IP industry are unprepared for the challenges posed by AI [13][14]. - The traditional barriers to creating infringing content, such as the need for technical skills and software, have been eliminated, making it accessible to anyone, including children [14][13]. - Menkes advocates for IP professionals to understand AI technology and anticipate the issues and opportunities it presents, as the scale and speed of potential infringements far exceed traditional monitoring capabilities [14][15]. Group 3: Recommendations for IP Holders - IP holders should assess their current copyright monitoring systems to ensure they can handle the scale of AI-generated content [14][15]. - Companies are encouraged to test new AI tools for built-in copyright safety mechanisms and to push for the implementation of filtering systems if such mechanisms are lacking [15][16]. - Establishing a rapid response mechanism to address infringements promptly is crucial, as the future of IP enforcement may depend on the ability to compel AI companies to implement protective measures [16][14]. Group 4: Future Outlook - There is concern that only large corporations will be able to effectively navigate the evolving copyright landscape, potentially leaving individual creators and small companies vulnerable to exploitation by AI [17][14]. - The article suggests that while AI may necessitate urgent updates to copyright laws, the outcome remains uncertain, with ongoing debates about the implications for creators and IP holders [17][14].

Core Viewpoint - Disney has accused Google of large-scale copyright infringement by using AI models and related services to generate and distribute images and videos without authorization, prompting Disney to send a cease-and-desist letter to Google [1][4]. Group 1: Legal Actions - Disney's lawyers have sent a letter to Google's Chief Legal Officer, claiming that Google has copied a significant amount of Disney's works as training material for its AI systems [1]. - The letter includes examples of AI-generated content that allegedly infringes on Disney's copyrights, specifically mentioning characters from popular films such as "Frozen," "The Lion King," and "Star Wars" [4]. Group 2: Demands from Disney - Disney demands that Google immediately cease the copying, displaying, distributing, or creating derivative content based on Disney characters across all Google AI output products, including YouTube and YouTube Shorts [4]. - The company insists that Google implement technical restrictions in its AI services to prevent the future generation of infringing content [4]. Group 3: Context of the Accusation - This accusation against Google follows similar cease-and-desist letters Disney has sent to Meta and Character.AI, and aligns with Disney's joint lawsuit with NBCUniversal and Warner Bros. Discovery against Midjourney and Minimax for copyright infringement [4]. - Disney argues that Google's market position is being strengthened by distributing AI services that exploit the notoriety of Disney's copyrighted works, causing further harm to Disney [4].

Core Insights - The emergence of generative AI has raised concerns about its potential to disrupt various industries, including education and intellectual property (IP) law, as even young children can inadvertently create copyright-infringing content using AI tools [1][9]. Group 1: AI and Copyright Issues - A six-year-old child was able to create a fully functional interactive website called "Bedtime Story Weaver" using Google AI Studio in under two minutes, without any coding experience [3][4]. - The website allows users to generate complete stories and illustrations based on simple inputs, highlighting how AI democratizes creative capabilities [3][4]. - The child generated a story featuring characters from different franchises, such as Sonic and Mario, which raises significant copyright infringement concerns [6][9]. Group 2: Industry Response and Recommendations - Jonathan Menkes, an IP lawyer, emphasized that the IP industry is unprepared for the challenges posed by AI, as the speed and scale of potential infringement far exceed traditional monitoring capabilities [9][10]. - He recommended that IP holders assess their copyright monitoring systems, test new AI tools for built-in copyright protections, and establish rapid response mechanisms to address infringement [10][11]. - Menkes also noted that companies like Disney are exploring user-generated content within controlled environments, suggesting that the future may favor brands that can balance IP protection with user creativity [11][12]. Group 3: Implications for Different Stakeholders - There is a concern that large companies will dominate copyright enforcement, while individual creators and small businesses may struggle to protect their IP against AI-generated content [12][13]. - The disparity in resources between major corporations and smaller entities could exacerbate existing inequalities in IP rights, as only large firms can afford legal defenses and negotiations with AI companies [13][14]. - The ongoing debates about copyright law and AI highlight the urgent need for updates to existing regulations, although the future landscape remains uncertain [13][14].

Core Argument - The U.S. Supreme Court is considering a case involving Cox Communications, which is seeking to avoid financial liability in a significant music copyright lawsuit brought by record labels accusing the company of enabling user piracy of thousands of songs [1][2]. Group 1: Legal Context - The justices expressed skepticism regarding Cox's claim that mere awareness of user piracy should not result in liability for copyright infringement [2][3]. - A previous jury found Cox liable for $1 billion due to secondary liability for infringement by its customers, which involved over 10,000 copyrights [5][7]. - The 4th U.S. Circuit Court of Appeals overturned the damages award in 2024, leading to a retrial to determine the amount owed to the labels [5][7]. Group 2: Implications for Internet Service Providers (ISPs) - ISPs are generally not held liable for user infringement if they take reasonable preventive measures, but the labels argue that Cox failed to address numerous infringement notices and did not cut off access for repeat infringers [6]. - The justices are concerned about the potential impact on innocent users if copyright enforcement becomes overly broad, which could lead to ISPs terminating service for entire households or institutions based on a single user's infringement [3][11]. Group 3: Industry Reactions - Major tech companies, including Alphabet's Google, Amazon, and Microsoft, have supported Cox in this case, while music, film, and book industry trade groups have backed the record labels [12].

Core Points - The article discusses the implications of generative AI in academic writing, particularly focusing on the issue of plagiarism and copyright infringement [1][2][3] - It emphasizes the distinction between plagiarism and copyright infringement, noting that while plagiarism is an ethical violation, copyright infringement is a legal issue [5][8][10] Group 1: Plagiarism and AI - Generative AI tools like ChatGPT are widely used in academic writing, with a significant percentage of students reporting their use for assignments [3] - The outputs from generative AI can create a false sense of originality, leading users to unknowingly present others' ideas as their own [4][16] - The lack of clear attribution in AI-generated content breaks traditional citation chains, complicating the identification of original sources [3][4] Group 2: Legal and Ethical Boundaries - Copyright laws generally prohibit the reproduction of creative expressions but do not protect ideas themselves, allowing for the sharing of thoughts without infringement [5][10] - The article highlights that generative AI outputs typically do not infringe copyright as they do not exhibit substantial similarity to the protected expressions used in training data [6][10] - There is a growing concern that the conflation of plagiarism and copyright infringement could lead to misunderstandings in legal contexts [7][10] Group 3: Distinction Between Concepts - Plagiarism is defined as the unauthorized use of another's language, ideas, or works without proper attribution, while copyright infringement involves the unauthorized use of protected expressions [9][10][13] - The article outlines that not all unethical academic behaviors constitute plagiarism, and some may not even infringe copyright [13][14] - The need for clear definitions and boundaries between copyright infringement, plagiarism, and poor academic practices is emphasized [8][11] Group 4: Attribution Rights - The article discusses the lack of universal attribution rights in U.S. law, suggesting that while attribution is important, it does not always constitute a legal violation [14] - Proposals for establishing new attribution rights are met with skepticism due to the complexity of copyright law and the potential for conflicting interpretations [14] - The importance of maintaining academic integrity and transparency in the use of AI-generated content is highlighted, advocating for clear guidelines in academic institutions [16]

Core Viewpoint - The recent court ruling against Visual China confirms the company's infringement of photographer Dai Jianfeng's rights, requiring an apology and compensation of 15,000 yuan [2][5][8]. Group 1: Court Ruling and Company Response - The Tianjin Peace District People's Court ruled that Visual China infringed on Dai Jianfeng's rights regarding the work "Village Under the Galaxy," requiring the company to publish a statement on its homepage and pay compensation [2][3][8]. - Visual China acknowledged the infringement, stating that the image was uploaded by a third-party contributor who violated copyright rules, and the company has since taken measures to remove the infringing content [9]. - The court found that Visual China's previous demand for compensation from Dai Jianfeng was inappropriate and lacked due diligence in rights verification [8]. Group 2: Financial Performance and Market Reaction - Visual China's stock price has seen a significant increase, with a rise of over 20% within the week, closing at 25.34 yuan and a total market capitalization of 17.753 billion yuan as of November 21 [9]. - For the third quarter of 2025, Visual China reported total revenue of 610 million yuan, a year-on-year increase of 0.30%, while net profit attributable to shareholders decreased by 9.03% to 74.314 million yuan [9].

Core Viewpoint - Visual China has publicly apologized for copyright infringement after a court ruling confirmed that the company unlawfully displayed and sold a photograph by photographer Dai Jianfeng, requiring compensation of 15,000 yuan [1] Group 1: Legal Proceedings - The Tianjin Peace District People's Court ruled that Visual China infringed upon Dai Jianfeng's rights regarding the photograph titled "Village Under the Galaxy," which was uploaded by a third party without authorization [1] - The court mandated Visual China to publish a statement on its official website to mitigate the impact of the infringement [1] Group 2: Company Response - Visual China has removed the infringing photograph from its website following the lawsuit and has ceased the infringing activities [1] - The company expressed its commitment to enhancing the protection of copyright holders' rights and improving its content review processes to contribute to a healthier copyright ecosystem [1]

Core Insights - OpenAI launched its advanced video generation model Sora 2 and the accompanying Sora App, which quickly became popular, topping the App Store in the U.S. and surpassing competitors like ChatGPT and Gemini [2][3][11] - The Sora App allows users to create immersive short videos by simply uploading a photo or entering a text prompt, marking a significant advancement in AI video generation technology [6][11] - The app's "Cameo" feature has led to a surge in user-generated content (UGC), with a reported 1200% increase in related content within 24 hours of its launch [11] Company Overview - OpenAI's valuation has reached $500 billion, and the company reported $4.3 billion in revenue for the first half of the year, exceeding its total revenue from the previous year, although it also faced a net loss of $13.5 billion [11] - The company is considering a potential IPO in the future and is currently facilitating a secondary sale of employee stock [11] Technology and Features - Sora 2 has made significant improvements over its predecessor, including synchronized audio and video generation, enhanced physical accuracy, and improved resolution and detail [6][11] - The app not only serves as an AI tool but also as a social entertainment platform where users can create and share video content [6][11] User Engagement - The introduction of the "Cameo" feature has made OpenAI's CEO Sam Altman a prominent figure on the internet, with users creatively incorporating his likeness into various scenarios [7][9] - Users in China are particularly active, creating localized content that features popular characters and trends [9] Legal and Copyright Concerns - The rapid rise of Sora has raised concerns about copyright infringement, as many videos generated using the app feature protected intellectual property (IP) without proper authorization [12][15] - OpenAI's "opt-out" policy for copyright content has been criticized for potentially increasing legal risks, as it allows users to generate content featuring copyrighted characters unless the rights holders request removal [15][16] - Legal experts warn that users could face multiple forms of infringement, including copyright and portrait rights, if they use the app to create videos featuring recognizable characters or individuals without permission [16][18]