Centralized cloud storage, like that offered by hyperscalers AWS S3, Google Cloud, and Microsoft Azure, were once promising alternatives to on-premise storage. But recently, companies are finding that centralized storage can be problematic due to issues like rising costs and complexities as data grows—plus increasing concerns around security and privacy.
‍
These are the hard facts creating mounting operational challenges for companies who need and rely on accessibility, performance, privacy and scale from their cloud storage solutions.
‍
While rising costs and security risks are motivating companies to look for alternatives, the most significant driver is a perceived lack of data privacy. Users are more aware of privacy in general and concerned about what personal data is collected and shared when they utilize a service or application. The largest centralized cloud storage providers have built their overall businesses on products that use personal data and activity to increase their revenue. At a time when large technology companies have been fined hundreds of millions of dollars for violating data privacy laws, organizations, developers and product teams are sometimes suspicious of the motivations of those same companies in storing their data.
‍
In addition, insider threats and negligence often rank as the #1 offender in terms of privacy and security breaches. For example, according to research from Verizon Business, 85% of breaches involve the human element. So, the more people involved in touching and managing your data, the more likely a potential breach. And the complexities and costs around implementing a Zero Trust implementation to avoid these risks are significant.
‍
Many companies have recognized these as valid concerns and are trying to show their users they’re doing all they can to protect their data privacy. A part of this effort is finding a viable alternative for centralized cloud storage to ensure data privacy. Decentralized cloud storage has evolved to a point where it can now be compared to the centralized storage model as not only a viable alternative, but one that delivers better privacy and security - with less cost and complexity. Keep reading to learn how centralized and decentralized storage services handle data privacy and the risks associated with both.

Data security is about making sure any data provided is protected from malicious hackers. Data privacy, on the other hand, is focused on metadata, the data about your data, and ensuring partners involved in an application—and that includes cloud storage providers—aren’t harvesting metadata.
‍
WhatsApp came under fire earlier this year as users of the secure messaging app on iPhones were shocked by a new iMessage privacy update. They found that WhatsApp was collecting much more metadata than competitors, and they were then harvesting the metadata and sharing it with Facebook. WhatsApp tried to make a privacy policy update back in January 2021, which resulted in millions of users abandoning the app for competitors like Signal.
‍

The reality is that WhatsApp data sharing has been happening with Facebook since 2016. Users generally don’t have a choice and are forced to accept current privacy policy changes or potentially lose access to the app entirely. Some may leave for competitors, some may stay, but the drama with WhatsApp and Facebook has raised two important issues:
‍
1. End-users are more aware than ever before about data privacy
2. Data privacy is important enough that significant revenue is won or lost over this issue

Developers have always been wary about data privacy, but even more so today. They’re more careful about what metadata they’re collecting—that said, metadata can be an important part of business value to understand user behavior. Understanding trends in how users utilize the application can help developers improve their products, and so some collection may be necessary. If a business decision is made to collect metadata, developers need to ensure it isn’t shared with partners who use it for their gain—especially if this sharing isn’t in line with your user’s wishes.
‍
Connecting the dots makes utilizing AWS, Google, and Microsoft cloud storage a perceived concern by end-users relative to their data privacy. Each of these companies has other business products relying on personal information, giving them the motivation to access the metadata stored on their service to enhance their product offerings. If you’ve got a great application, end-users are willing to provide personal data in order to use it. But they aren’t so keen on that information being used for other purposes—particularly for revenue gain.
‍
Despite what privacy policies might say, developers are now worried that users may perceive a lack of data privacy when using these storage providers. And as seen in the recent WhatsApp example, this poses a significant risk to their revenue model.

Cloud storage applies to any service in which data is transmitted and stored on remote storage systems. Centralized and decentralized storage have significant differences in their architecture and design, and these architectural differences have an enormous impact on data privacy.
‍
This illustration represents how data is transferred and stored in centralized vs. decentralized storage systems, and it’s a stark difference.

The basic model of centralized cloud storage is built on an infrastructure of virtualized storage, where cloud-based data gets stored in logical pools across commodity storage servers. These centralized servers are in data centers located in various regions throughout the world to provide availability. Efficiencies are gained by metering these resources within the data centers, thereby making it cost-effective for data storage.
‍
Most centralized cloud storage providers replicate storage across multiple data centers—they do this to ensure data is available should a region experience a natural disaster, power outage, or malicious attack, which would mean the data being stored in that location wouldn’t be available. Take, for instance, AWS—they have three zones of replication, and in theory, if one zone should go down, the other two should be able to keep data available for their customers. Cloud storage data centers do their best to employ security measures to keep data protected. Still, ultimately, the model involves “honeypots” of data that are centralized within physical buildings—WikiLeaks has even made these locations public knowledge.
‍
Centralized cloud storage providers also offer data encryption, but this is always at an additional cost and is not included in the base offer. The real problem with this is that unless you’re willing to pay extra for client-side encryption, these providers can access your data whenever they want.

Decentralized cloud storage is a collection of services and responsibilities brought together under a zero-trust architecture to ensure data privacy and security. Instead of data centers, decentralized storage utilizes shared hard drives worldwide, referred to as Storage Nodes.
‍
Storage Nodes are computers, servers, and other storage devices run by individuals or companies loaning out unused space who get paid for their storage capacity and bandwidth. All of the data stored on Storage Nodes are default encrypted and erasure-coded. This means that in the decentralized model, the people hosting the data can’t ever access it.
‍
Decentralized cloud storage also includes network management services to monitor the health of the overall network and ensure data integrity. These services can be centralized or distributed depending upon the model. One of the most secure decentralized cloud storage providers uses Satellite uplinks to store and retrieve data across Storage Nodes. The uplink client automatically encrypts object data and metadata.
‍
Data encryption is a standard part of decentralized cloud data storage as it is necessary to maintain the zero-trust security posture, where the system is designed assuming no entity that interacts with the data can be trusted.