{
  "id": 20836426,
  "title": "GDPR Article 22's \"right to an explanation\" is an engineering requirement in disguise",
  "url": "https://urgent.news/2026/11/20/gdpr-article-22s-right-to-an-explanation-is-an-engineering",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-11-20T01:44:54.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/sithan/gdpr-article-22s-right-to-an-explanation-is-an-engineering-requirement-in-disguise-5hmo"
  },
  "original_language": "en",
  "account": "Every time Article 22 comes up, lawyers debate the meaning of \"meaningful information about the logic involved,\" while engineers rush to SHAP plots. However, the true difficulty lies not in the model being a black box, but in the inability to locate the specific decision in question. The right to an explanation under Article 22 requires that you can locate the decision, reconstruct its inputs, and provide a human-readable explanation. This involves retrieving the decision for that particular person on that date, ensuring the inputs are accurate, and explaining the outcome in simple terms. The explanation must be understandable without technical jargon and must be recorded in a tamper-proof manner. Teams often struggle to locate the relevant data, and without a proper audit trail, the process becomes futile. The core issue is not technical but organizational — ensuring each automated decision is treated as a first-class, immutable event, with inputs and explanations recorded and preserved. The challenge is retrofitting auditability into existing systems, which can be a significant undertaking. To test compliance, pick an automated decision from last week and determine if you can produce all necessary information for a regulator's review. If you can, the system is on the right path.",
  "summary": "Every time Article 22 comes up, the same argument starts. Lawyers debate what \"meaningful information about the logic involved\" is supposed to mean, and engineers start reaching for SHAP plots. I think both camps are overcomplicating it. I've been on the receiving end of this requirement — fintech, hiring tools, the usual suspects — and what actually gets you in trouble is never \"the model is a…",
  "key_points": [
    "Article 22 requires explanation of automated decisions",
    "Finding specific decision and inputs is challenging",
    "Auditability and organization are core compliance issues"
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}