01
Data isolation framework (core mechanism)
YOCO applies classification and access control across accounts, works, references, templates, intermediates, and logs.
- 账号与机构授权范围校验
- 作品和文件访问权限控制
- Enterprise-Grade专属实例部署能力
- Skill Token有效期与撤销机制
Enterprise-Grade Security
YOCO Security and Compliance
After a user starts a task, YOCO creates a PPT outline, slide content, and narration script from topic, requirements, and references, then applies templates to produce the deck. References are only processed within the scope required for that task. Outputs can be previewed, edited, regenerated, and only downloaded after confirmation.
1) Data Security and Privacy
End-to-end security across storage, transmission, and processing
02
Encryption and secure transport
Data is transmitted through secure protocols between users, platform, and third parties with strict file/API access controls.
- 安全通信协议
- 数据库与对象存储访问控制
- 文件预签名与权限校验
- 密钥和Token不写入业务日志
Data use scope (important)
YOCO uses data only within the necessary scope: [Processing scope] - User input topics, prompts, and generation settings - User uploaded reference materials - AI-generated PPT outline, per-slide content, and scripts - Presentation structure parsing (slide / element / layout) - Animation timeline parsing - Multimedia sync processing (audio/video / 3D models) [Usage boundaries] - Reference parsing is only for the current task and is not written into a knowledge base - User inputs, references, and outputs are not used to train platform base models - Not shared with any other user task - Not used for ad targeting or user profiling - Only required operational data is transmitted to essential third-party services
II. Content security, copyright and accountability
Content risk is controlled through automated rules, user confirmation, random manual checks, and complaint handling.
01
Content ownership structure
All user-uploaded content remains the IP of the user or authorized rights holder.
- Prompts and references belong to user or rights holder
- Presentation outputs belong to user
- Documents/images/videos belong to user or original rights holder
- 3D models and media belong to the original rights holder
02
Generated content and usage responsibilities
AI-generated PPTs and scripts are assistive outputs based on user instructions and materials and should be reviewed and confirmed before use.
- 生成内容支持用户编辑
- 不保证生成内容完全准确
- 不改变参考资料的原始版权归属
- 由用户确认后使用与分发
Liability boundary (legal separation principle)
User responsibilities: - - Content legality - - Copyright provenance - - Compliance of usage behavior YOCO responsibilities: - Provide AI content generation and editing - Provide content safety governance and structural checks - Provide structure parsing capability - Provide video generation capability - Provide secure execution environment - Handle complaints and take actions based on verification results
Content safety governance model
Automatic mechanisms: - Third-party foundation model safety capabilities - Input file, business parameter, and output structure validation - Prompt restrictions, exception retries, and task state control Human and user mechanisms: - User preview, edit, and regenerate - Periodic manual sampling of completed tasks - Complaint review, abnormal high-frequency calls, and high-risk task audits Review boundaries: - The platform does not manually review every single generated item - Structural checks are not full semantic review
III. Generated content labeling and user rights
Enable users and recipients to identify AI-generated content and trace its source.
01
Explicit labels
Prompt users during generation and download that content includes AI-generated elements for visibility.
- 生成界面提示内容由AI生成
- 下载时再次提示AI生成属性
- PPT最后一页显示“YOCO AI生成”
- 标识不遮挡PPT主要内容
02
Implicit labeling and traceability
Write traceability metadata in PPTX properties and custom fields without affecting rendering.
- 记录AIGC属性与生成工具
- 关联PPT作品唯一编号
- 记录生成时间和工具版本
- 隐式标识不因取消Explicit labels而删除
User awareness and control
Users can: - View outline, slide previews, scripts, and task status - Edit, regenerate, save, or delete generated results - Confirm again before download that the content includes AI-generated elements - Submit complaints, corrections, or deletion requests via feedback/support channels 重要提示： - AI-generated outputs may contain factual errors or omissions; please verify before use - Do not use AI outputs directly as professional conclusions for healthcare, legal, finance, public policy, or news content
IV. System processing principles (core logic)
All generation and conversion are initiated by users; outputs are released after confirmation.
Processing model details
YOCO processing logic: 1) Input layer - User submits topic, prompts, and generation settings - Users can upload reference files - Users can also upload existing PPTs and media 2) Generation and confirmation layer - Generate PPT outline, per-slide content, and scripts - Run safety checks on inputs and generated content - Users review and update output 3) File and video output layer - Generate editable PPTX from confirmed content - Parse existing slide pages, animations, and media structure - Sync scripts, narration, subtitles, and avatars - Produce final video after user confirmation Core principles: - Minimum necessary processing - Outputs are reviewable and editable - Export after user confirmation and labeling
Minimum necessary processing
Reviewable and editable outputs
Export after confirmation and labeling
One-line summary
YOCO processes only the data needed for each user request. Users confirm outputs, and the platform does not create public feeds or publish content on behalf of users.