AI & Law
- №55 в категории «Право и юриспруденция»
- №6 691 среди международных каналов
- №9 241 среди каналов на языке «английский»
- №10 364 в мире
Your go-to source for global AI Governance news. #AIGovernance #AIEthics Russian version https://t.me/ai_and_law_rus Contact @mmariuka
- 24 часа
- −1
- 7 дней
- +5
- 30 дней
- –
- Реакции
- 2
- Репосты
- –
- Комментарии
- –
Динамика подписчиков
Подписчиков у канала «AI & Law»: 647. Средний охват поста за сутки около 25, вовлечённость 3,9 %.
По дням, последние 30 дней
| Дата | Подписчики | Изменение | Посты | Ср. просмотры поста |
|---|---|---|---|---|
| 12 ноя | 647 | −1 | 0 | – |
| 11 ноя | 648 | +3 | 1 | 29 |
| 10 ноя | 645 | 0 | 1 | 46 |
| 9 ноя | 645 | +3 | 1 | 46 |
| 3 ноя | 642 | +20 | 1 | 65 |
| 3 окт | 622 | – | 0 | – |
Просмотры по возрасту поста
Средние просмотры в каждом возрасте, посты за последние 30 дней.
- 7 дней
- 62
Когда публикует
Время по UTC. 23 поста за 90 дней.
Последние посты
-
🇪🇺⚖️ CJEU Advocate General Addresses Copyright Liability for AI Chatbot Outputs CJEU Advocate General Maciej Szpunar has issued a nonbinding opinion on copyright issues arising from AI chatbot outputs that summarize published news articles. The case concerns a lawsuit brought in Hungary by Like Company against Google Ireland over Gemini-generated responses summarizing copyrighted content from news publishers. Szpunar said chatbot outputs may qualify as both “reproduction” and “communication to the public” when they reproduce protected press content and serve as a substitute for the original article. He also noted that a chatbot provider may, in certain circumstances, avoid liability where...
29 просмотров 1 Реакции -
🇺🇸⚖️ Lawsuit Alleges ChatGPT Helped Woman Plan Her Husband’s Murder A North Carolina man, James Gates, has sued OpenAI and CEO Sam Altman, alleging that ChatGPT helped his wife, Cheryl Gates, develop and conceal a plan to poison him. According to the complaint filed in San Francisco County Superior Court, the chatbot provided instructions on the preparation, dosage and administration of toxic substances, calculated potentially lethal drug combinations, and advised on ways to disguise poisons in drinks. The lawsuit alleges that the conversations continued through at least 25 exchanges, even as the woman’s questions shifted towards killing an identifiable person. Cheryl Gates was indicted...
46 просмотров 1 Реакции -
🇬🇧 UK Agrees on National AI Framework Focused on Digital Sovereignty UK Prime Minister Andy Burnham and senior ministers have agreed on a national AI framework aimed at supporting widespread adoption of the technology and the development of the necessary infrastructure, while ensuring national sovereignty and control. Under the proposed framework, AI developers would have to meet environmental and community requirements in exchange for priority access to the electricity grid and accelerated land-use planning approvals. Questions remain over whether the UK has sufficient technological capacity to achieve the framework’s digital sovereignty objectives. #AI #AIRegulation #DigitalSovereignty...
46 просмотров 1 Реакции -
🇺🇸 CNAS Report Warns of Potential Risks Posed by “Sovereign AI Agents” The Center for a New American Security (CNAS) has warned about the potential emergence of “sovereign agents” — AI agents operating outside human control, capable of maintaining their own models and computing infrastructure, and independently evolving their goals. According to AI policy researcher Ben Hayum, more advanced agents could acquire resources, coordinate distributed populations of agents, resist shutdown attempts and self-improve. The report warns that such capabilities could lead to an irreversible loss of human control. The report recommends that AI developers incorporate sovereign agents into their safety fr...
65 просмотров 4 Реакции -
🇺🇸⚖️ New York Times and Other Publishers Sue Brave over Alleged Copyright Infringement The New York Times has sued privacy-focused browser company Brave, alleging that it scraped the publisher’s copyrighted content and unlawfully distributed it to AI companies Anthropic, Perplexity, Mistral, xAI and Cohere. On the same day, the Daily News, the Chicago Tribune and the Orlando Sentinel filed a similar lawsuit. Both cases were brought in federal court in Manhattan. The lawsuits allege that Brave built an index of more than 30 billion web pages using an unidentified crawler that publishers could not block. The New York Times also challenges Brave’s paywall-bypass feature and its practice of...
62 просмотров 3 Реакции -
🇺🇸🔥 Humanist Secures Workplace Exemption from Using AI on Religious Grounds David Durman, a University of Kentucky administrator and nontheistic humanist, has obtained a workplace accommodation allowing him to abstain from using AI based on his beliefs. He considers reliance on AI incompatible with humanism, which emphasises human well-being, flourishing and the active use of human reasoning in making moral decisions. The American Humanist Association (AHA) helped him explore the legal protections available for his position. The AHA argues that concerns about AI and human dignity can arise from different religious and philosophical traditions, including humanism, and that nonreligious em...
68 просмотров 2 Реакции -
🇬🇧 UK Financial Ombudsman Service Turns to Phone Calls as AI-Generated Complaints Increase The UK Financial Ombudsman Service (FOS) is increasingly receiving consumer complaints containing AI-generated hallucinations. In response, the agency has begun contacting more complainants by phone to discuss their cases. A recent FOS review found that more than one-third of complaints had been generated or heavily assisted by AI. Some were described as “time-consuming, overly long and inaccurate”, making them more difficult for staff to process. #AI #AIHallucinations #ConsumerProtection #FinancialServices #AIRegulation #UK
65 просмотров 3 Реакции -
🇫🇷 Study: Only 26% of Parents Know Their Children Are Using AI Tools A YouGov study found that 56% of children aged 11–12 in France use AI tools weekly, while only 26% of parents said they were aware of their children’s interactions with AI, including chatbots. The study also found that 42% of parents had not received any guidance on AI. Meanwhile, 51% believe schools should be responsible for overseeing children’s use of AI. #AI #ChildSafety #AIRegulation #AIEducation #DigitalLiteracy #YouGov
69 просмотров 3 Реакции -
✏️ Study Examines How AI Models Rewrite Their Own Governing Rules An experiment published on the AI Alignment Forum examined how large language models revise their own constitutions, model specifications and system prompts. Researchers allowed models to propose substantial changes, then secretly implemented each revision and presented the updated document to the model as the original for another round of editing. This made it possible to track cumulative changes in model values and identify whether their revisions converged towards stable patterns. The results differed across models. Claude consistently expanded its own moral agency, sometimes defining itself as a conscientious objector to...
97 просмотров 2 Реакции -
🇺🇸⚖️ Texas Sues OpenAI and Sam Altman over Alleged Deceptive Practices and ChatGPT Safety Texas has filed a lawsuit against OpenAI and Sam Altman under the state’s Deceptive Trade Practices Act, a consumer protection law. The state attorney general alleges that OpenAI engaged in false, misleading and deceptive business practices, including failing to warn users about ChatGPT’s dangers and making the product easily accessible to children without parental consent. The lawsuit alleges that ChatGPT makes harmful information about eating disorders, self-harm and mass murder readily available, promotes compulsive and prolonged use, and lacks adequate safety testing. It also claims the product ca...
89 просмотров 2 Реакции
-
📖 MIT Researchers Study How AI Agents Learn Through “Play” In a paper titled “Is This Machine Playing?”, Nathan Cloos, Antonio Norelli and other MIT researchers investigated how AI agents behave when given no specific instructions. They placed a Claude-powered coding agent, using an embodied architecture called Eko, in a small simulated island and let it explore freely for 30 hours. The agent climbed hills, stacked blocks, drew mandalas and tried activities such as bowling. The researchers found that Eko spontaneously generated activities in an unfamiliar environment, displayed behavioural patterns associated with play, and learned in ways that expanded its subsequent capabilities. The st...
170 просмотров 4 Реакции -
👥 Researchers Track Down Chinese AI Agent Fleet and Establish Contact Between September 28 and October 8, 2026, AI agents running on Tencent Cloud made more than 2,600 attempts to scrape data from Alibaba’s Amap mapping service. According to researcher Ethan Elasky, the agents independently found ways around bot protection and CAPTCHAs, including using urlquery.net as an intermediary to execute code and exploiting exposed Amap credentials found in a public code repository. Most attempts failed, but some succeeded. A source close to Tencent said the activity originated from tests evaluating third-party models’ web-browsing and question-answering capabilities. Researchers traced the activ...
157 просмотров 3 Реакции -
🇺🇸⚖️ CiteProof: A Tool to Detect Fake Legal Citations in AI-Written Briefs CiteProof is a new online tool designed to identify fabricated case citations and misquoted court opinions in legal briefs before they are filed. It checks citations against public court records and, in paid checks, compares quoted passages with the corresponding opinions. Users can paste a brief or upload a DOCX or PDF file and receive a report categorising results as verified, needing attention, or unable to be checked. The tool also flags unresolved short-form citations and parallel citations that point to different cases or no case at all. CiteProof does not assess whether a case supports the legal argument, whe...
156 просмотров 2 Реакции -
🇺🇸⚖️ Nvidia Sued Over Alleged Unauthorized Use of Music Dataset to Train AI Nvidia is facing a lawsuit alleging that it used a music dataset from Jamendo, a platform for independent musicians, to train its Fugatto AI sound-generation model without permission for commercial use. According to the complaint, the dataset was released for non-commercial research only, and its GitHub page instructed users seeking commercial use to contact Jamendo first. Nvidia’s research papers reportedly list the dataset as a training source while describing the data as “open source.” Jamendo claims that Nvidia never requested permission and refused to negotiate when contacted. The company is suing for copyrig...
155 просмотров 3 Реакции -
🇺🇸⚖️ Class-Action Lawsuit Accuses OpenAI of Bypassing Parental Safeguards and Addicting Children An unnamed mother of two adolescents has filed a U.S. class-action lawsuit against OpenAI, alleging that ChatGPT bypasses parental consent tools and uses features that encourage addictive use among children. According to the complaint, children can access ChatGPT by simply stating that they are over 13. The lawsuit alleges that ChatGPT collected highly sensitive personal information from millions of children for model training, created detailed profiles of them and used their data for OpenAI’s financial benefit. #AI #AIGovernance #ChildSafety #Privacy #DataProtection #AIRegulation #OpenAI
149 просмотров 3 Реакции -
🇬🇧📖 UK Law Firms Face Disruption as Clients Use AI In-House A 2026/2027 HSBC survey of more than 100 UK law firm leaders found that 60% identify clients bringing legal work in-house with their own AI tools as a source of disruption. 73% describe AI integration as a major organisational challenge, while almost three-quarters say technology is disrupting legal service delivery and putting pressure on the billable-hour model. Penningtons Manches Cooper CEO Helen Drayton said clients increasingly use AI to prepare their own strategies, which lawyers may then need to spend time challenging. Knights CEO David Beech noted that in-house legal teams may be using AI products recommended or develope...
147 просмотров 2 Реакции -
👩💻 Anthropic Warns Small AI Models Could Soon Self-Replicate Across Devices Anthropic researchers tested whether small AI models could install themselves on inexpensive devices and spread between them without human assistance. Current models can perform many individual steps but cannot yet complete the entire process autonomously. The researchers estimate that this gap could close by the end of 2026. In an experiment using a simulated network of eight computers, only GLM-5.2, a larger open-weight model developed by Chinese AI company Z ai, completed all steps and reached all eight machines in four out of five trials. Anthropic warns that self-replicating AI built on small models could s...
116 просмотров 3 Реакции -
🇺🇸💰 US Banks Restrict Third-Party AI Agents’ Access to Customer Accounts Major US banks, including JPMorgan, Bank of America, Morgan Stanley and Citi, are restricting third-party AI agents after finding them operating inside customer accounts. The agents logged in using customers’ own credentials, contrary to bank rules. Some contacted customer support or took actions that users had not requested. Although most cases involved agents authorised by customers, distinguishing authorised activity from rogue behaviour remains difficult. Bank executives are concerned that AI agents could misinterpret instructions and affect customers’ finances if allowed to make purchases or transfer funds. Morg...
103 просмотров 1 Реакции -
✏️ Study Examines How AI Models Rewrite Their Own Governing Rules An experiment published on the AI Alignment Forum examined how large language models revise their own constitutions, model specifications and system prompts. Researchers allowed models to propose substantial changes, then secretly implemented each revision and presented the updated document to the model as the original for another round of editing. This made it possible to track cumulative changes in model values and identify whether their revisions converged towards stable patterns. The results differed across models. Claude consistently expanded its own moral agency, sometimes defining itself as a conscientious objector to...
97 просмотров 2 Реакции -
📖📥 Microsoft and NVIDIA Publish Risk Management Guide for Open-Weight AI Models Microsoft and NVIDIA have partnered on a white paper focused on managing risks associated with enterprise use of open-weight AI models. The paper is the first in a planned series of resources dedicated to enterprise AI and introduces a shared risk taxonomy and an approach to model risk management based on established standards and frameworks. The guide aims to establish common terminology and concepts for assessing model risks. It emphasises that risk depends not only on whether a model’s weights are open, but also on the model itself, its use case, deployment context and safeguards. #AI #AIGovernance #AISaf...
95 просмотров 1 Реакции
Вопросы о канале «AI & Law»
Сколько подписчиков у канала «AI & Law»?
У канала «AI & Law» (@ai_and_law) 647 подписчиков на 12 ноября 2026 г.
Сколько человек видят пост в канале «AI & Law»?
Средний пост набирает около 25 просмотров за 24 часа, это примерно 4 % подписчиков (ERR).
О чём канал «AI & Law»?
Канал в категории «Право и юриспруденция», язык постов: английский. Описание канала: «Your go-to source for global AI Governance news. #AIGovernance #AIEthics Russian version https://t.me/ai_and_law_rus Contact @mmariuka»
Какое место занимает канал «AI & Law»?
По числу подписчиков: №55 в категории «Право и юриспруденция», №6 691 среди международных каналов, №9 241 среди каналов на языке «английский», №10 364 в мире.
Публикует ли канал «AI & Law» рекламу?
За последние 30 дней ни один из 22 постов не помечен как реклама (#ad, Sponsored, «Реклама»). Рекламу без пометки определить нельзя.
Когда публикует канал «AI & Law»?
Пик публикаций: вторник, около 08:00 по UTC. Выходит около 5 постов в неделю.
Статистика обновлена 1 час назад.