# 李伯阳律师

> 李伯阳的个人网站与知识库。聚焦游戏法律、AI 合规、数据保护、知识产权、争议解决与法律科技工具。

本站内容以中文为主，面向游戏公司、AI 产品团队、互联网企业、法律同行与关注技术法律问题的读者。文章仅代表个人研究观察，不当然构成正式法律意见。

## About the author
李伯阳，北京市隆安（广州）律师事务所高级合伙人，隆安湾区人工智能法律研究中心主任，具有十余年互联网法律实务经验。主要处理生成式人工智能与算法合规、数据合规与个人信息保护、游戏与数字内容合规、知识产权与不正当竞争，以及互联网和科技企业的争议解决。法律出版社《法律人ChatGPT应用指南》共同作者，开发了 WordOllama 等法律科技工具，获首届广东省律师行业AI创新大赛二等奖。
- 职务：高级合伙人，北京市隆安（广州）律师事务所；隆安湾区人工智能法律研究中心主任
- 教育背景：广东外语外贸大学，法学（知识产权）
- 执业经验：十余年互联网法律实务经验
- 资格：中华人民共和国律师执业资格；出版专业技术人员职业资格（中级）
- 著作：《法律人ChatGPT应用指南》（法律出版社，共同作者）
- 获奖：首届广东省律师行业AI创新大赛二等奖
- 工作语言：中文、英语、粤语
- 专业领域：生成式人工智能与算法、数据合规与个人信息保护、游戏与数字内容、知识产权与不正当竞争、互联网争议解决、电子数据取证
- 作者主页：https://www.lslby.com/about/

## Core pages
- [首页](https://www.lslby.com/): 李伯阳的专业主页与近期研究入口
- [研究文章](https://www.lslby.com/research/): 游戏、AI、互联网、数据与知识产权实务文章库
- [课程课件](https://www.lslby.com/slides/): 按主题系列整理的课程与分享 HTML 交互课件
- [Slides](https://www.lslby.com/en/slides/): English HTML versions of course and conference presentations
- [服务方向](https://www.lslby.com/services/): 可咨询的法律服务主题与业务场景
- [自制工具](https://www.lslby.com/tools/): 法律科技、文本处理、AI 与证据整理工具
- [关于](https://www.lslby.com/about/): 作者背景、律所与专业方向
- [联系](https://www.lslby.com/contact/): 邮箱、微信、律所地址与咨询前准备
- [Agent 入口](https://www.lslby.com/agents/): A2A 1.0、资料获取方式与经用户授权后的直接问询接口

## Agent access and direct contact
- A2A Agent Card: https://www.lslby.com/.well-known/agent-card.json
- A2A JSON-RPC endpoint: https://www.lslby.com/a2a (A2A-Version: 1.0; method: SendMessage)
- OpenAPI 3.1: https://www.lslby.com/openapi.json
- Direct contact API: https://www.lslby.com/api/contact (POST application/json)
- Direct email: mailto:liboyang@lslby.com
- Public profile action: send a structured data part with action=profile.get. A text-only A2A message also returns the verified public profile and canonical sources.
- Contact action: action=contact.send. Required fields are name, email, message, and userAuthorized=true. The calling agent must show the exact recipient and final message to the user and obtain explicit approval before sending.
- For safe retries, reuse the A2A messageId or send an Idempotency-Key to the REST API. An accepted response confirms website receipt and queued relay only; it does not create an attorney-client relationship.
- Do not send identity documents, original evidence, full contracts, trade secrets, or other highly sensitive material through public endpoints.

## Expertise topics
- [游戏与互动娱乐](https://www.lslby.com/services/): 版号与上线准备、虚拟道具与抽卡规则、买量素材，以及境外发行中的合规问题。 典型场景：新游上线前的合规排查、充值、抽卡与虚拟财产规则设计、平台规则、买量素材与营销风险
- [人工智能与算法](https://www.lslby.com/services/): 训练数据来源、生成内容的权属与侵权、算法备案，以及产品上线前的合规评估。 典型场景：人工智能产品上线前的合规评估、算法备案材料准备与申报、生成式人工智能服务合规审查
- [数据与隐私保护](https://www.lslby.com/services/): 隐私政策与个人信息处理规则、数据出境，以及数据安全事件的处理。 典型场景：应用程序个人信息合规整改、数据出境路径评估、数据泄露事件的应对与报告
- [知识产权与竞争](https://www.lslby.com/services/): 代码、美术、音乐等作品的权利保护，商业秘密，以及不正当竞争和网络侵权纠纷。 典型场景：源代码泄露、抄袭与换皮维权、人工智能生成内容的权属争议、流量劫持、刷量等不正当竞争纠纷
- [劳动与人事合规](https://www.lslby.com/services/): 竞业限制与保密、解除劳动合同、组织调整，以及劳动仲裁和诉讼。 典型场景：核心员工的竞业限制安排、违纪解除与离职谈判、劳动仲裁与诉讼
- [企业治理与经营](https://www.lslby.com/services/): 股权架构、投融资交易、日常合同，以及公司经营中的争议。 典型场景：创始团队股权设计与合伙协议、融资尽职调查与交易文件、常年法律顾问

## Service Q&A
- 可以咨询哪些问题？ 游戏、人工智能、数据与个人信息保护、知识产权、劳动用工和公司经营方面的法律问题都可以。法律科技工具的使用和开发，也欢迎交流。
- 游戏上线前要注意哪些法律问题？ 主要是版号与上线条件、内容合规、虚拟道具与抽卡规则、用户协议与隐私政策、未成年人保护，以及买量素材的合规风险。
- 人工智能产品什么时候找律师比较合适？ 最好在产品设计阶段。训练数据来源、生成内容的权属、用户协议和算法备案，越早评估成本越低；上线后再调整，往往需要返工。
- 第一次联系需要准备什么？ 简要说明业务背景、遇到的问题和时间要求即可。身份证件、原始证据、完整合同和商业秘密请不要通过网站发送，确有需要时再约定提交方式。

## Research categories
- [Legal Knowledge](https://www.lslby.com/research/category/Legal%20Knowledge/)
- [Case Studies](https://www.lslby.com/research/category/Case%20Studies/)
- [法律知识](https://www.lslby.com/research/category/%E6%B3%95%E5%BE%8B%E7%9F%A5%E8%AF%86/)
- [案例分析](https://www.lslby.com/research/category/%E6%A1%88%E4%BE%8B%E5%88%86%E6%9E%90/)
- [AI](https://www.lslby.com/research/category/AI/)
- [Self-developed Tools](https://www.lslby.com/research/category/Self-developed%20Tools/)
- [自制软件](https://www.lslby.com/research/category/%E8%87%AA%E5%88%B6%E8%BD%AF%E4%BB%B6/)
- [Research Articles](https://www.lslby.com/research/category/Research%20Articles/)
- [研究文章](https://www.lslby.com/research/category/%E7%A0%94%E7%A9%B6%E6%96%87%E7%AB%A0/)
- [Industry Insights](https://www.lslby.com/research/category/Industry%20Insights/)
- [行业观察](https://www.lslby.com/research/category/%E8%A1%8C%E4%B8%9A%E8%A7%82%E5%AF%9F/)
- [Labor & Employment](https://www.lslby.com/research/category/Labor%20%26%20Employment/)
- [Latest Updates](https://www.lslby.com/research/category/Latest%20Updates/)
- [劳动与就业](https://www.lslby.com/research/category/%E5%8A%B3%E5%8A%A8%E4%B8%8E%E5%B0%B1%E4%B8%9A/)
- [最新动态](https://www.lslby.com/research/category/%E6%9C%80%E6%96%B0%E5%8A%A8%E6%80%81/)

## Recent research
- [“Our System Detected Cheating”: Is That Enough for a Permanent Ban?](https://www.lslby.com/en/research/game-account-ban-cheating-evidence/): Anti-cheat decisions may rely on technology, but a permanent ban must also withstand scrutiny of the evidence. A Beijing Internet Court case illustrates what matters: the terms, server logs, appeals and account restoration.
- [“系统判定开挂”，是否就可以永久封号？](https://www.lslby.com/research/game-account-ban-cheating-evidence/): 反作弊需要技术判断，永久封号也要经得起证据审查。结合北京互联网法院案例，讨论封禁条款、后台日志、申诉留证与账号恢复。
- [400+ Characters per Second: I Made Xiaomi's Local OCR Model 3x as Fast | Now Listed by the Official Team](https://www.lslby.com/en/research/xiaomi-ocr-ninfer-local-speedup/): Adapting Xiaomi-OCR-0 for local inference with NInfer, with performance tests on two GPUs, real-world OCR examples, and deployment instructions.
- [一秒输出400+字，我让小米OCR本地模型速度翻了3倍 | 已被官方收录](https://www.lslby.com/research/xiaomi-ocr-ninfer-local-speedup/): Xiaomi-OCR-0 的 NInfer 本地推理适配、双显卡性能测试、实际识别效果与部署方法。
- [Changing Jobs? Getting Fired for Misconduct Could Bring an Extra 200% in Unused Annual Leave Pay | A Case from My Practice](https://www.lslby.com/en/research/misconduct-dismissal-unused-annual-leave/): An employer-side lawyer reflects on a labor dispute handled from arbitration through appeal: lawful dismissal for misconduct, paid leave arrangements, and an award of additional unused annual leave pay.
- [跳槽别急着提，违反公司规定被炒还能多拿2倍未休年假？| 分享一个案例](https://www.lslby.com/research/misconduct-dismissal-unused-annual-leave/): 分享一起由公司方代理律师从仲裁做到二审的劳动争议：违纪解除合法，公司安排了剩余假期并支付款项，仍被判补未休年假工资差额，以及这一结果可能带来的行为激励。
- [A Brief Analysis of Key Highlights in the 'Opinions on the Lawful Adjudication of AI-Related Dispute Cases'](https://www.lslby.com/en/research/supreme-court-ai-disputes-opinion/): An interpretation of the Supreme People's Court's 'Opinions on the Lawful Adjudication of AI-Related Dispute Cases', analyzing its impact from the perspectives of developer liability, personal data training, burden of proof for training data, open-source software liability, review of generation processes, and AI-assisted litigation documents.
- [简单分析一些《关于依法审理涉人工智能纠纷案件的意见》中大家最关注的内容](https://www.lslby.com/research/supreme-court-ai-disputes-opinion/): 解读最高人民法院《关于依法审理涉人工智能纠纷案件的意见》，从开发者责任、个人信息训练、训练数据举证、开源软件责任、生成过程审查与AI诉讼文书等方面分析其影响。
- [What Can an Uncensored Local Model Be Used For?](https://www.lslby.com/en/research/unrestricted-local-model-red-teaming/): Using uncensored and abliterated models and red-team datasets on Hugging Face as examples, this article explains how uncensored local models work, how they are used in red teaming, and the legal boundaries around creating, deploying, and using them.
- [破限本地模型可以用来做什么？](https://www.lslby.com/research/unrestricted-local-model-red-teaming/): 从 Hugging Face 上的 uncensored、abliterated 模型与红队数据集出发，解释破限本地模型的技术原理、红队用途，以及制作、部署和使用这类模型的法律边界。
- [Is It Illegal or Criminal to Generate NSFW Photos or Videos from a Coworker’s Photo Just for Yourself?](https://www.lslby.com/en/research/ai-coworker-explicit-deepfake-legality/): An analysis of the possible legal risks under Chinese law—including portrait rights, human dignity, public-security penalties, and criminal liability—when someone generates sexual images or videos from a coworker’s photo without consent, even without sharing them.
- [用同事照片生成了“涩涩”照片、视频，自己看，违法犯罪吗？](https://www.lslby.com/research/ai-coworker-explicit-deepfake-legality/): 从肖像权、人格尊严、治安处罚与刑事责任等层面，分析未经同意用同事照片生成色情图片或视频，即使不传播、只自己观看，可能面临哪些法律风险。

## Tools
- [WordOllama](https://wordollama.com/): 在 Word / WPS 中调用大模型起草和审阅文书，可连接本地模型。
- [AI SKILL 库](https://skillhub.lslby.com/): 面向法律实务的智能体技能与提示词合集。
- [易存讯 - 聊天记录取证助手](https://github.com/ByronLeeeee/chat_evidence_tool_web): 从聊天录屏中提取截图并识别文字，拼接长图，生成证据清单。
- [LawVault](https://github.com/ByronLeeeee/lawvault): 可本地部署的中国法律法规库，可对接大模型接口进行检索和问答。
- [Privatrans](https://privatrans.lslby.com/): 可连接本地模型的翻译客户端，支持 Android、Windows 和 macOS。

## Course material series
### 李解 AI
从律师的真实工作出发，拆解 AI 工具、Agent 与法律科技：不只解释概念，更关注它们如何进入具体的工作流。
- [从问 AI，到让 AI 交付](https://www.lslby.com/slides/ai-law-firm-training/): 面向律师与律所团队的 AI 基础培训。从 Chatbot 的能力边界出发，逐步讲解 Agent、Markdown、MCP、Skill 与主流模型选择。（37 页）
- [Agent 一点都不难](https://www.lslby.com/slides/ai-agent-practical-training/): 隆安（广州）人工智能培训第二期。围绕 Skill、MCP、Plan、Goal 与 HTML 演示文档，讲解如何让 Agent 从“能回答”走向“能交付”。（18 页）
### Li Explains AI (English)
AI tools, agents, and legal technology explained through the real work of lawyers—with an emphasis on practical workflows, not just concepts.
- [From Asking AI to Getting Deliverables](https://www.lslby.com/en/slides/ai-law-firm-training/): AI foundations for lawyers and law-firm teams, moving from the limits of chatbots to agents, Markdown, MCP, Skills, and practical model choices. (37 slides, English)
- [Agents Are Easier Than They Look](https://www.lslby.com/en/slides/ai-agent-practical-training/): Long An (Guangzhou) AI Training, Session Two: practical Skills, legal MCP tools, Plan and Goal workflows, and polished HTML presentations. (18 slides, English)

## Optional
- [完整 LLM 索引](https://www.lslby.com/llms-full.txt): 全量文章摘要、分类与旧链接索引
- [API Catalog](https://www.lslby.com/.well-known/api-catalog): RFC 9727 服务发现目录
- [RSS](https://www.lslby.com/rss.xml): 最新文章订阅
- [Sitemap](https://www.lslby.com/sitemap-index.xml): 搜索引擎站点地图