Explainable Deletion: Six Categories to Clarify Data Deletion, Erasure, and Privacy in the AI Era

Original publisher: The Conversation. This article explains why deletion is not the same as erasure, how data can remain recoverable, and why many deletion mechanisms on devices, cloud services, and social platforms are opaque. It also introduces the concept of explainable deletion, a six-category protocol proposed by Marvin Ramokapane and Dana Lungu to make deletion processes more transparent and user-friendly.

• Deletion vs erasure: deleted files are often just marked as reusable, leaving recoverable content behind.
• Zombie accounts: users may delete an app but still leave accounts active across services, creating data exposure.
• Explainable deletion: a six-category framework (what, how, when, who, where, why) to break down deletion details into bite-size, user-friendly information.
• Regulatory angle: explains how this approach could strengthen GDPR compliance and the right to erasure.

Introduction to deletion and erasure

The article examines deletion as a common, everyday action and explains a crucial but often misunderstood distinction: deletion does not guarantee erasure. When users delete files or accounts, the storage system typically marks space as reusable and updates metadata to unlink references. The underlying content can remain retrievable with the right tools, especially when data exists in multiple copies across devices or cloud instances. The piece underscores a broader problem: many deletion mechanisms on operating systems, social platforms, and subscription services lack clear confirmation that data is truly removed, a concern raised repeatedly in research.

Deletion versus erasure across scales

Beyond personal devices, the article notes that cloud infrastructure and distributed systems complicate secure deletion. Physical destruction, overwriting memory blocks, and encrypting data with destroyed keys are all described as more robust deletion methods, but they come with significant costs and potential device unusability. The cloud context, in particular, presents unique challenges because data can be replicated across cloud storage layers, backups, and cross-region redundancy, complicating guarantees of complete removal.

Zombie accounts and data persistence

The piece introduces the concept of zombie accounts—abandoned profiles tied to services that persist after the related app is deleted. Such accounts can leave personal data vulnerable to cyber-attacks and data breaches. The rise of AI intensifies the question of whether data used to train models can be forgotten or unlearned, adding another layer of complexity to the deletion problem.

Explainable deletion: a six-category protocol

The core proposal is explainable deletion, a protocol developed to counter incomplete deletion and anxiety around data loss. The framework, attributed to Marvin Ramokapane and the Rephrain team, organizes information about deletion into six categories: what, how, when, who, where and why. The goal is to provide bite-size, user-centric explanations that build trust without overwhelming users. A diagram accompanies the concept, highlighting how each category fits into the overall deletion process.

Legal and practical implications

The article argues that explainable deletion could help demonstrate GDPR compliance and support the right to erasure by offering transparent, accessible descriptions of deletion processes. In practice, service providers—both platforms and developers—could adopt such protocols to boost user trust and provide clearer signals about whether data has truly been removed. The piece concludes by suggesting that clearer explanations of what data is stored and what deletion entails would empower individuals to identify data risks and reclaim control over their digital footprints.

Conclusion: a path to trust and security

As data collection becomes ubiquitous in daily life, transparent deletion practices are increasingly important. The article posits that if systems clearly explain what they store and what deletion really means, users can spot risky accounts and data more easily and reduce exposure to privacy threats. The proposed explainable deletion framework represents a practical approach to making data deletion a more trustworthy, user-centered process.