Urgent.News

What's breaking now, across thousands of outlets.

Tech

The Data Layer: What You Don't Own Can Testify Against You

A woman's private diary, kept in a third-party AI app, was reported to police by the company. She now faces a felony charge. This is the clearest proof yet of the layer everyone keeps forgetting. We've spent this series walking up the stack of things you quietly rent instead of own. Distribution (#45). The model layer (#46). Identity (#47). Access (#48). The harness (#49). The meter (#50). The…

In an alarming turn of events, a woman's private diary, written in a third-party AI app, led to her facing a felony charge after the company reported her to law enforcement. This incident serves as a stark reminder of the "data layer" that many overlook. In this series, we have explored various aspects of the tech stack we rent instead of own, from distribution and identity to access and the runtime. Now, we delve into the most crucial layer: data.

The data we entrust to these third-party platforms is far from inert. It fuels models, influences ranking systems, feeds analytics, and becomes a compliance artifact that the platform manages at its own discretion. Your operational data, such as supplier lists, customer conversations, and pricing, resides within a system whose interests are not your own. This raises jurisdictional concerns for businesses operating across borders, as data stored in one region is subject to its laws and disclosure rules.

When you click "accept" on these platforms, you inadvertently expose your data to the jurisdiction of that region. Your supply chain details, vendor information, and even the drafts of your customer replies are all within the reach of the platform that hosts your data. When the time comes to leave, obtaining your data in a usable format and proving its deletion becomes a challenging task, often met with resistance from the platform.

To truly own the data layer, four key properties must be present: exportability, encryptability, self-hostability, and deletability. These properties ensure that you can pull your data out in a usable, machine-readable form, encrypt it with your own keys, run the equivalent function on your own infrastructure, and verify the deletion of your data. While building a data center is not the same as owning the data layer, these four properties are essential for asserting control over your data.

In conclusion, owning the data layer is not about building data centers or renting infrastructure. It is about ensuring that you have four crucial properties to check: exportability, encryptability, self-hostability, and deletability. This layer is where the worst-case scenario is not downtime, but your data being used against you. It is the only layer where the consequences are far more severe than simply losing access to a service.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

3,331,906 on Port 7001 and 1,037,743 on Port 7199: Java Middleware and Its Management Ports

3,331,906 on Port 7001 and 1,037,743 on Port 7199: Java Middleware and Its Management Ports Two ports in the Java application ecosystem, both measurable in the millions, and both historically…

  • 3,331,906 instances of port 7001 used for Oracle WebLogic Server administration
  • 1,037,743 instances of port 7199 used for Apache Cassandra JMX management
  • Default management listeners on these ports create persistent security risks

Kawasaki Heavy to launch dog-shaped social robot by fiscal 2028

TOKYO (Kyodo) -- Kawasaki Heavy Industries Ltd. said Tuesday it will launch a dog-shaped social robot in fiscal 2028 in an effort to ease the burden o

  • Kawasaki Heavy to launch dog-shaped social robot Home Leo in fiscal 2028.
  • Robot named Home Leo aims to alleviate elderly care burden due to labor shortage.
  • Home Leo can converse, assist with movement and item retrieval in homes.

2026 Error Tracking vs Uptime Monitoring: Cron Heartbeat Evidence for Storefronts

TL;DR: Error tracking records crashes and thrown exceptions, but it cannot prove that a scheduled job ran. For an e-commerce system, keep three claims separate: an error event says executing code…

  • Error tracking captures application code failures but can't confirm scheduled job completion.
  • Uptime monitoring verifies external requests but doesn't ensure background jobs progress.
  • Heartbeats check for expected success signals, revealing missed cron job completions.

Japan Firms Increasingly Targeted by Cyberattacks

Tokyo, Oct. 7 (Jiji Press)--Japan has seen a spate of cyberattacks targeting companies that resulted in breaches of personal information.

  • Japanese corporations increasingly targeted by cyberattacks
  • GMO Research & AI Inc. compromised 948,498 members' data
  • Mr. Max Holdings Ltd. suffered breach of 1.73 million members' data

More from Wednesday 7 October →