vendor-ai-review

Review AI vendor contracts for data usage, responsibility, and compliance risks.

183|37|Updated May 15, 2026
One-click install
npx skills add https://github.com/zhou210712/claude-for-legal-ZH --skill vendor-ai-review-zhou210712
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: vendor-ai-review
Source: https://github.com/zhou210712/claude-for-legal-ZH/tree/main/ai-governance-legal/skills/vendor-ai-review
Command: npx skills add https://github.com/zhou210712/claude-for-legal-ZH --skill vendor-ai-review-zhou210712

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

解决AI供应商合同中常见的合规与责任风险不清晰问题,帮你系统核查训练数据合规性、责任分配、模型变更通知与合规义务向下传导是否到位,并输出可执行的谈判立场与红线条款。

Core Features & Use Cases

  • 训练数据来源与用途核查:重点检查供应商是否将客户数据用于训练、是否区分个人信息/敏感个人信息、是否提供合法来源与删除承诺等。
  • AI特有责任分配审查:识别供应商将AI侵权、内容安全与错误损失风险单方面转嫁给客户的条款模式,并给出理想与最低接受标准。
  • 模型变更与合规传导核查:检查模型变更通知机制、性能退化补救与终止权安排,以及算法备案、安全评估等合规义务是否能有效落实并可审计。
  • 输出可直接用于谈判的结构化结果:给出风险汇总、红线标记、谈判优先级与下一步决策树(经核准/附条件核准/阻止/升级法律顾问)。

Quick Start

使用 vendor-ai-review 审查你要签署的AI SaaS/API/模型授权合同条款,并将需要审查的供应商文本粘贴到参数中。

Frequently Asked Questions about vendor-ai-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review AI vendor contracts for training data compliance risks?

AI vendor contract review identifies ambiguous data usage and responsibility risks by checking training data sourcing, responsibility allocation, model change notification, and downstream regulatory compliance propagation to produce risk summaries and redline negotiation positions.

What AI specific liability risks should I look for in SaaS or API agreements?

AI vendor contract review identifies single-sided liability transfers for AI infringement, content safety, and output errors by evaluating responsibility allocation clauses, generating structured risk summaries with ideal and minimum acceptance standards for negotiation positions.

How do I check model change notification mechanisms in AI model licensing contracts?

Checking model change notification mechanisms in AI model licensing involves verifying if the vendor provides update alerts, handles performance degradation, and grants termination rights. This review assesses these governance terms and checks if algorithm filing and safety assessments are auditable.

Can I use this to assess regulatory compliance propagation for AI services?

Yes, you can assess regulatory compliance propagation by checking if algorithm filing, safety evaluations, and downstream compliance obligations are effectively allocated and auditable within AI vendor contracts, producing a decision tree for conditional approval or escalation.

What is the best way to prepare redline negotiation positions for AI vendor terms?

The best way to prepare redline negotiation positions is to map identified risks across data usage, liability, and model governance into a structured priority list, which this skill outputs directly alongside risk summaries and an approval decision tree.