本文へ移動
cccskills
無料GitHub で公開

pia-vendor-processing

Conducts Privacy Impact Assessment for vendor and third-party data processing arrangements. Covers processor due diligence, Data Processing Agreement (DPA) requirements under GDPR Article 28, sub-processor management, cross-border vendor transfers, cloud service provider assessments, and ongoing vendor monitoring. Keywords: vendor PIA, processor assessment, DPA, Article 28, sub-processor, cloud privacy, third-party risk.

インストール方法を見る

含まれるファイル(5)

  • SKILL.md7.7 KB
  • assets/template.md3.1 KB
  • references/standards.md2.7 KB
  • references/workflows.md3.9 KB
  • scripts/process.py14.6 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Privacy Impact Assessment for Vendor Processing

Overview

When a controller engages a processor (vendor) to process personal data on its behalf, GDPR Article 28 imposes specific obligations on the controller to ensure that the processor provides sufficient guarantees to implement appropriate technical and organisational measures. The EDPB has repeatedly emphasized that the controller remains accountable for the processing regardless of delegation to a vendor. This skill provides a structured PIA methodology for assessing and managing the privacy risks of vendor data processing arrangements.

Regulatory Framework

GDPR Article 28 — Processor Obligations

RequirementDescription
Art. 28(1)Controller shall use only processors providing sufficient guarantees
Art. 28(2)Processor shall not engage another processor (sub-processor) without prior specific or general written authorisation
Art. 28(3)Processing governed by a contract or legal act setting out subject matter, duration, nature, purpose, data types, categories of data subjects, and controller obligations/rights
Art. 28(3)(a)-(h)Mandatory DPA clauses: process only on documented instructions, confidentiality, security measures, sub-processor conditions, assist with data subject rights, assist with security/breach/DPIA, delete or return data, provide audit information
Art. 28(4)Sub-processor must be bound by the same data protection obligations
Art. 28(5)Adherence to approved codes of conduct or certification (Art. 42) as elements to demonstrate sufficient guarantees

Processor Selection Criteria (EDPB Guidance)

The controller must assess the processor's:

  • Technical expertise and security measures
  • Track record and reputation (including enforcement history)
  • Ability to ensure data subject rights can be exercised
  • Data location and cross-border transfer implications
  • Sub-processor chain management capability
  • Business continuity and data recovery capability
  • Willingness to submit to audits

Vendor Risk Categories

Risk CategoryExamplesRisk Level
Cloud infrastructure (IaaS/PaaS)AWS, Azure, GCP — hosting personal dataHigh (data at rest and in transit under vendor control)
SaaS applicationsCRM, HRIS, email marketing — processing personal data in vendor applicationHigh (vendor application logic and access controls)
Analytics providersWebsite analytics, marketing attribution — processing behavioural dataMedium-High (data sharing, profiling, cookie compliance)
IT support / managed servicesHelpdesk, system administration — access to personal data during supportMedium (access-based risk, not purpose-based)
Payment processorsCard processing, billing — financial personal dataHigh (PCI DSS intersection, financial data sensitivity)
Sub-processorsVendor's own vendors processing controller's dataHigh (reduced visibility and control)

PIA Methodology for Vendor Processing

Phase 1: Vendor Inventory and Classification (Week 1)

  1. Inventory all vendors processing personal data on behalf of the organisation.
  2. For each vendor, document: personal data categories processed, data subject categories, processing purposes, data volumes, data locations.
  3. Classify each vendor by risk category based on data sensitivity, processing scope, and access level.
  4. Identify the vendor's sub-processors and their roles.
  5. Determine cross-border transfer implications for each vendor.

Phase 2: Due Diligence Assessment (Week 2)

  1. Request and review the vendor's security certifications (ISO 27001, SOC 2, CSA STAR).
  2. Review the vendor's privacy certifications and GDPR compliance evidence (ISO 27701, APEC CBPR).
  3. Check the vendor's enforcement history with supervisory authorities.
  4. Assess the vendor's technical and organisational security measures against the controller's requirements.
  5. Evaluate the vendor's data breach notification capability and track record.
  6. Review the vendor's sub-processor management process.
  7. Assess the vendor's ability to assist with data subject rights requests.

Phase 3: Data Processing Agreement Review (Week 3)

  1. Verify the DPA covers all Art. 28(3) mandatory provisions.
  2. Check that processing instructions are specific and documented (not generic catch-all clauses).
  3. Verify sub-processor authorisation mechanism: specific or general with notification and objection rights.
  4. Ensure data deletion or return obligations at contract termination are clear and enforceable.
  5. Verify audit rights: right to conduct or commission audits of the processor.
  6. Check cross-border transfer mechanisms (SCCs, BCRs, adequacy) for international vendors.
  7. Assess liability and indemnification provisions for data protection breaches.

Phase 4: Risk Assessment and Mitigation (Week 4)

  1. For each vendor, assess the risk to data subjects from the vendor's processing.
  2. Consider: data breach at vendor, unauthorised access by vendor employees, sub-processor chain risk, vendor lock-in, business continuity failure.
  3. Score risks on likelihood and severity matrix.
  4. Identify mitigation measures: encryption (controller-managed keys), pseudonymisation before transfer to vendor, access restrictions, monitoring, contractual controls.
  5. Assess residual risk after mitigation.

Phase 5: Ongoing Monitoring Framework (Week 5)

  1. Define vendor review frequency based on risk classification (high-risk: annual; medium: biennial).
  2. Establish vendor security questionnaire process for periodic reassessment.
  3. Monitor vendor sub-processor changes (notification process, assessment of new sub-processors).
  4. Track vendor incidents and breach notifications.
  5. Conduct or commission periodic audits for high-risk vendors.
  6. Define vendor exit strategy: data return, deletion verification, transition plan.

Enforcement Precedents

  • CJEU Wirtschaftsakademie (C-210/16, 2018): Facebook fan page administrator held to be joint controller with Facebook; controllers cannot avoid responsibility by using a vendor platform.
  • CJEU Fashion ID (C-40/17, 2019): Website operator embedding Facebook Like button held to be joint controller for the collection and transmission of data; responsibility for vendor components on own website.
  • Belgian DPA vs IAB Europe (2022): EUR 250,000 fine; TCF framework operator held to be controller, not processor; demonstrates that contractual designation does not determine controller/processor status.
  • Spanish AEPD vs Vodafone (2021): EUR 8.15 million fine for failure to adequately manage processors and sub-processors handling customer data; insufficient DPA controls.
  • Finnish DPA vs Posti Group (2020): EUR 100,000 fine for inadequate processor management; controller failed to verify processor's compliance with data processing instructions.
  • Dutch DPA guidance on cloud services (2022): Detailed assessment methodology for cloud vendor privacy compliance, including requirement to conduct DPIA for cloud migrations involving personal data.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Implements 42 CFR Part 2 protections for substance use disorder patient records. Covers written consent requirements stricter than HIPAA, re-disclosure prohibition, court order procedures, qualified service organization agreements, and 2024 amendments aligning Part 2 with HIPAA. Keywords: 42 CFR Part 2, substance use disorder, SUD records, re-disclosure, consent, Part 2 amendments.

日本語の概要は準備中です。原文の説明を表示しています。

mukul975/Privacy-Data-Protection-Skills3032026年3月17日 更新

Guides assessment of third-country adequacy decisions under GDPR Article 45 for international data transfers. Covers the current EC adequacy decisions list, adequacy assessment criteria, partial adequacy handling, and monitoring of adequacy decision reviews. Keywords: adequacy decision, Article 45, third country, adequate protection, EC adequacy list.

日本語の概要は準備中です。原文の説明を表示しています。

mukul975/Privacy-Data-Protection-Skills3032026年3月17日 更新

Implements age-gating mechanisms for online services to restrict access based on user age. Covers hard gates versus soft gates, neutral age prompts, re-verification triggers, circumvention prevention, and regulatory requirements under GDPR, COPPA, UK Online Safety Act, and DSA. Keywords: age gate, age restriction, neutral prompt, children, online services, access control.

日本語の概要は準備中です。原文の説明を表示しています。

mukul975/Privacy-Data-Protection-Skills3032026年3月17日 更新

Evaluates and implements age estimation and verification technologies for online services. Covers facial age estimation, digital ID verification, self-declaration with risk assessment, AI-based age estimation, and the accuracy versus privacy tradeoff. Includes ICO guidance and euCONSENT framework. Keywords: age verification, age estimation, facial analysis, digital ID, children, online safety.

日本語の概要は準備中です。原文の説明を表示しています。

mukul975/Privacy-Data-Protection-Skills3032026年3月17日 更新

Preparing EU AI Act compliance documentation for high-risk AI systems. Covers Annex III classification, technical documentation under Art. 11, conformity assessment, risk management systems, and CE marking requirements. Keywords: EU AI Act, high-risk AI, Annex III, conformity assessment, CE marking.

日本語の概要は準備中です。原文の説明を表示しています。

mukul975/Privacy-Data-Protection-Skills3032026年3月17日 更新

Implements GDPR Art. 22 automated decision-making and AI Act Art. 14 human oversight requirements for AI systems. Covers identification of solely automated decisions, meaningful human intervention design, logic explanation mechanisms, and contestation procedures. Keywords: Art. 22, automated decision, human oversight, AI Act, profiling, contestation.

日本語の概要は準備中です。原文の説明を表示しています。

mukul975/Privacy-Data-Protection-Skills3032026年3月17日 更新

mukul975 のスキルをすべて見る

このスキルの問題を報告する